<?xml version="1.0" encoding="UTF-8"?><rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/" xmlns:atom="http://www.w3.org/2005/Atom" version="2.0" xmlns:cc="http://cyber.law.harvard.edu/rss/creativeCommonsRssModule.html">
    <channel>
        <title><![CDATA[Stories by Padmesh Sharma on Medium]]></title>
        <description><![CDATA[Stories by Padmesh Sharma on Medium]]></description>
        <link>https://medium.com/@padmesh2224?source=rss-64f4d52198ee------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/1*_oq8O6bU4a0noOtwc2No0g.jpeg</url>
            <title>Stories by Padmesh Sharma on Medium</title>
            <link>https://medium.com/@padmesh2224?source=rss-64f4d52198ee------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Sun, 22 Nov 2026 22:40:26 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@padmesh2224/feed" rel="self" type="application/rss+xml"/>
        <webMaster><![CDATA[yourfriends@medium.com]]></webMaster>
        <atom:link href="http://medium.superfeedr.com/" rel="hub"/>
        <item>
            <title><![CDATA[Docker Tutorial for Beginners: A Step-by-Step Guide]]></title>
            <link>https://medium.com/@padmesh2224/docker-tutorial-for-beginners-a-step-by-step-guide-52547b47fec4?source=rss-64f4d52198ee------2</link>
            <guid isPermaLink="false">https://medium.com/p/52547b47fec4</guid>
            <category><![CDATA[docker-compose]]></category>
            <category><![CDATA[docker]]></category>
            <category><![CDATA[dockerfiles]]></category>
            <category><![CDATA[docker-image]]></category>
            <category><![CDATA[containerization]]></category>
            <dc:creator><![CDATA[Padmesh Sharma]]></dc:creator>
            <pubDate>Sat, 13 Sep 2025 12:20:22 GMT</pubDate>
            <atom:updated>2025-09-13T12:20:22.669Z</atom:updated>
            <content:encoded><![CDATA[<p>Tired of the classic “but it works on my machine” problem? Whether you’re a developer, designer, or data scientist, Docker is the game-changer you need. Think of it as a magic box 🪄 that bundles your code, dependencies, and environment into a single, portable unit that runs consistently anywhere.</p><p>In this step-by-step guide, we’ll demystify Docker and get you from zero to hero. We’ll cover:<br> ✔ What is Docker? (And why you should care)<br> ✔ How to install Docker on your machine<br> ✔ Essential Docker commands with practical examples<br> ✔ Building your own custom image with a Dockerfile<br> ✔ Orchestrating multi-container apps with Docker Compose<br> ✔ Best practices for writing efficient and secure containers</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*5GQwZvDKl2E9QwVV973ZGw.png" /></figure><h3>📌 What is Docker? Beyond the “Magic Box”</h3><p>At its core, Docker is a platform that uses containerization to make it easier to create, deploy, and run applications.</p><p><strong>Containers vs. Virtual Machines:</strong> Unlike VMs, which virtualize an entire operating system, containers virtualize at the application level. They share the host system’s kernel but run in isolated user spaces. This makes them incredibly lightweight, fast, and efficient.</p><p><strong>The Analogy:</strong> If a Docker Image is a blueprint or a recipe, a Docker Container is the actual, running house or meal made from that recipe.</p><h3>🔑 Key Concepts You Must Know</h3><p>Before diving into commands, let’s clear up some building blocks of Docker:</p><p><strong>1. Docker Image</strong></p><ul><li>A <strong>blueprint or recipe</strong>.</li><li>Contains everything your app needs: code, runtime, libraries, configurations.</li><li>Example: A Node.js image comes with Node.js already set up.</li></ul><p><strong>2. Docker Container</strong></p><ul><li>A <strong>running instance of an image</strong>.</li><li>If an image is the recipe, a container is the actual meal.</li><li>You can run multiple containers from the same image.</li></ul><p><strong>3. Docker Engine</strong></p><ul><li>The <strong>core service</strong> that runs Docker on your system.</li><li>Responsible for creating and managing containers and images.</li></ul><p><strong>4. Docker Hub</strong></p><ul><li>A <strong>registry (like GitHub for Docker)</strong> where you can find and share images.</li><li>Example: You can pull a ready-made Redis or MySQL image instead of building your own.</li></ul><p><strong>5. Dockerfile</strong></p><ul><li>A <strong>text file with instructions</strong> on how to build your own image.</li><li>Example: “Start from Node.js base image, copy my code, install dependencies, run the app.”</li></ul><p><strong>6. Docker Compose</strong></p><ul><li>A tool for managing <strong>multi-container apps</strong>.</li><li>Example: A web app + database can be started together with one command.</li></ul><h3>🤔 Why Use Docker?</h3><ul><li><strong>Consistency</strong> → Same environment everywhere, no more “works on my machine.”</li><li><strong>Portability</strong> → Runs on macOS, Windows, Linux, or any cloud.</li><li><strong>Efficiency</strong> → Lightweight compared to VMs.</li><li><strong>Isolation</strong> → Apps don’t clash with each other.</li><li><strong>Speed</strong> → Start and stop in seconds.</li></ul><h3>⚙️ Installing Docker</h3><ol><li>Head to the official Docker Desktop download page: <a href="https://www.docker.com/products/docker-desktop/">https://www.docker.com/products/docker-desktop/</a></li><li>Download and install the version for your operating system (Windows, macOS, or Linux).</li><li>Launch Docker Desktop. (It may require a restart).</li><li>Verify your installation by opening a terminal and running:</li></ol><pre>docker version<br># Example output: Docker version 24.0.2, build 1113bc5</pre><p>If you see a version number, congratulations! You’re all set! 🎉</p><h3>🛠 Docker Basics: Essential Commands Cheat Sheet</h3><ol><li><strong>Pulling an Image from Docker Hub</strong><br>Docker Hub is a public registry where you can find pre-built images. The pull command downloads them to your local machine.</li></ol><pre>docker pull ubuntu:22.04<br># Pro Tip: Always specify a tag (like :22.04) for reproducibility.</pre><p>Downloads an Ubuntu image from Docker Hub.</p><p><strong>2. Running Your First Container</strong><br> The run command creates and starts a container from an image.</p><pre># Run an interactive Ubuntu shell<br>docker run -it ubuntu:22.04 /bin/bash<br># -it = interactive terminal<br># Once inside, type `exit` to stop the container.</pre><p>Opens an interactive Ubuntu shell inside a container.</p><p><strong>3. Listing Containers &amp; Images</strong></p><pre>docker ps        # Shows only running containers<br>docker ps -a     # Shows all containers (including stopped ones)<br>docker images    # Lists all images downloaded on your system</pre><p><strong>4. Building Your Own Image with a Dockerfile</strong></p><p>Step 1: Create a simple Node.js app (index.js)</p><pre>const http = require(&#39;http&#39;);<br>const server = http.createServer((req, res) =&gt; {<br> res.writeHead(200, {&#39;Content-Type&#39;: &#39;text/plain&#39;});<br> res.end(&#39;Hello from my first Docker container!\n&#39;);<br>});<br>server.listen(3000, &#39;0.0.0.0&#39;, () =&gt; {<br> console.log(&#39;Server running on port 3000&#39;);<br>});</pre><p>Step 2: Create a Dockerfile</p><pre>FROM node:18-alpine<br>WORKDIR /app<br>COPY package*.json ./<br>RUN npm install<br>COPY . .<br>EXPOSE 3000<br>CMD [&quot;node&quot;, &quot;index.js&quot;]</pre><p>Step 3: Build and run your container</p><pre>docker build -t my-node-app:1.0 .<br>docker run -d -p 3000:3000 --name my-running-app my-node-app:1.0</pre><p>Visit <a href="http://localhost:3000/">http://localhost:3000</a> in your browser to see your app running!</p><p><strong>5. Stopping &amp; Cleaning Up</strong></p><pre>docker stop my-running-app<br>docker rm my-running-app<br>docker rmi my-node-app:1.0</pre><p><strong>6. Sharing Images via Docker Hub</strong></p><pre>docker login<br>docker tag my-node-app:1.0 your-dockerhub-username/my-node-app:1.0<br>docker push your-dockerhub-username/my-node-app:1.0</pre><p><strong>7. Managing Multi-Container Apps with Docker Compose</strong></p><p>Example docker-compose.yml for a Node.js app with Redis:</p><pre>version: &quot;3.8&quot;<br>services:<br> web:<br>   build: .<br>   ports:<br>     - &quot;3000:3000&quot;<br>   depends_on:<br>     - redis<br> redis:<br>   image: &quot;redis:alpine&quot;</pre><p>Run your stack:</p><pre>docker-compose up -d<br>docker-compose down</pre><h3>💡 Best Practices for Beginners</h3><p>➢ Use .dockerignore to exclude unnecessary files (like node_modules, .git).<br>➢ Choose small base images like -alpine to reduce image size.<br>➢ Leverage layer caching by copying dependencies first in your Dockerfile.<br>➢ Never store secrets in images; use environment variables or Docker. secrets.<br>➢ Use specific version tags instead of latest.</p><h3>📚 Further Learning &amp; Resources</h3><ul><li>Official Docker Documentation → <a href="https://docs.docker.com/">https://docs.docker.com/</a></li><li>Docker for Beginners (freeCodeCamp, YouTube)</li></ul><h3>💡 Wrapping Up</h3><p>You’ve taken your first steps into the world of containerization! You now know what Docker is, how to run containers, build your own images, and even manage multi-service apps. The best way to learn is by doing, so pick a small project and try to “Dockerize” it.</p><p>Got questions or stuck somewhere? Drop them in the comments below! 👇<br> If you found this guide helpful, please give it a clap 👏 — it helps others find it too.</p><p><strong>Padmesh Sharma</strong><br> 🔗 LinkedIn: <a href="https://www.linkedin.com/in/padmesh-sharma/">Padmesh Sharma</a><br> 📅 Book a 1:1 Chat: <a href="https://topmate.io/padmesh_sharma">topmate.io/padmesh_sharma</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=52547b47fec4" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Git Tutorial for Beginners: A Step-by-Step Guide]]></title>
            <link>https://medium.com/@padmesh2224/git-tutorial-for-beginners-a-step-by-step-guide-fbe502385f98?source=rss-64f4d52198ee------2</link>
            <guid isPermaLink="false">https://medium.com/p/fbe502385f98</guid>
            <category><![CDATA[git-commands]]></category>
            <category><![CDATA[tutorial]]></category>
            <category><![CDATA[git]]></category>
            <category><![CDATA[github]]></category>
            <category><![CDATA[version-control]]></category>
            <dc:creator><![CDATA[Padmesh Sharma]]></dc:creator>
            <pubDate>Fri, 16 May 2025 11:50:51 GMT</pubDate>
            <atom:updated>2025-05-16T11:50:51.129Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*cF-FG5Wr-5E2omlXGg-mpg.png" /></figure><p>Whether you’re a developer, designer, or data scientist, <strong>Git</strong> is an essential tool for managing your projects. It acts like a <strong>time machine for your code</strong>, tracking every change so you can collaborate seamlessly, undo mistakes, and experiment fearlessly.</p><p>In this guide, we’ll cover:<br>✔ <strong>What Git is and why it matters</strong><br>✔ <strong>How to install Git</strong><br>✔ <strong>Essential Git commands with examples</strong><br>✔ <strong>Best practices for version control</strong><br>✔ <strong>Working with remote repositories (GitHub/GitLab/Bitbucket)</strong></p><h3>📌 What is Git?</h3><p>Git is a <strong>version control system (VCS)</strong> that:</p><p>✅ Tracks every change in your codebase<br>✅ Allows seamless collaboration without conflicts<br>✅ Lets you revert to previous versions<br>✅ Supports branching for parallel development</p><p><strong>Without Git:</strong></p><ul><li>Accidentally deleting code? No way to recover.</li><li>Team collaboration? Chaos.</li><li>Experimenting with new features? Risky.</li></ul><p><strong>With Git:</strong></p><ul><li>Every change is recorded.</li><li>Multiple people can work together smoothly.</li><li>You can test ideas safely in branches.</li></ul><h3>🤔 Why Use Git?</h3><ol><li><strong>Version Control: </strong>Never lose your work, revert to any past state.</li><li><strong>Collaboration: </strong>Work with teams without overwriting changes.</li><li><strong>Backup &amp; Recovery: </strong>Mistakes happen, Git helps you undo them.</li><li><strong>Branching: </strong>Test new features without breaking the main code.</li></ol><h3>⚙️ Installing Git</h3><p>Install Git on your machine from the official site:</p><p>👉 <a href="https://git-scm.com/downloads">https://git-scm.com/downloads</a></p><p>Once installed, open <strong>Git Bash</strong> (Windows) or Terminal (Mac/Linux) and type:</p><pre>git --version</pre><p>If it returns a version, you’re all set!</p><h3>🛠 <strong>Git Basics: Essential Commands</strong></h3><h3>1. 🔧 Setting Up Git</h3><p>Configure your identity (used in commits):</p><pre>git config --global user.name &quot;Your Name&quot;  <br>git config --global user.email &quot;your@email.com&quot;</pre><p>💡 <em>Need help with configuration?</em></p><pre>git config --help</pre><h3>2. 🆕 Starting a Repository</h3><p>To initialize Git in a project folder:</p><pre>git init</pre><p>📁 This creates a .git directory where Git tracks your project history.</p><h3>3. ✍️ Tracking Changes</h3><p><strong>Stage Files:</strong></p><pre>git add .           # Stage all files  <br>git add file.txt    # Stage a specific file</pre><p><strong>Commit Files:</strong></p><pre>git commit -m &quot;Initial commit&quot;</pre><p>💬 <strong>Commit Best Practices:</strong><br> ✅ Use present tense (“Add feature”, not “Added feature”)<br> ✅ Keep messages clear and descriptive<br> ✅ Commit small, logical changes<br> ✅ Reference issues or task IDs when applicable</p><h3>4. 🔍 Checking Status &amp; History</h3><p><strong>View File Status:</strong></p><pre>git status          # See changes</pre><p><strong>View Commit History:</strong></p><pre>git log             # Full history  <br>git log --oneline   # Concise view  </pre><h3>5. ♻️ Undoing Changes</h3><p><strong>Unstage a File:</strong></p><pre>git restore --staged file.txt</pre><p><strong>Reset to a Previous Commit:</strong></p><pre>git reset --soft HEAD~1   # Keeps changes staged  <br>git reset --hard HEAD~1   # Discards all changes</pre><p>⚠️ <strong>Warning</strong>: git reset --hard deletes changes permanently. Make sure you don’t need them!</p><h3>6. 🌐 Working with Remote Repositories</h3><p><strong>Clone a Repo:</strong></p><pre>git clone https://github.com/user/repo.git</pre><p><strong>Push Changes:</strong></p><pre>git push origin main</pre><p><strong>Pull Updates:</strong></p><pre>git pull origin main</pre><h3>7. 🌿 Branching &amp; Merging</h3><p><strong>Create &amp; Switch Branches:</strong></p><pre>git branch new-feature      # 🌱 Create branch  <br>git checkout new-feature    # 🔄 Switch branch</pre><p><strong>Merge Branches:</strong></p><pre>git checkout main  <br>git merge new-feature</pre><h3>8. 📥 Stashing Changes</h3><p><strong>Save Uncommitted Changes Temporarily:</strong></p><pre>git stash          # Save changes</pre><p><strong>Restore Stashed Work:</strong></p><pre>git stash pop      # Restore changes</pre><p><strong>Clear Stash List:</strong></p><pre>git stash clear      # Remove changes</pre><h3>9. 🚫 Ignoring Files with .gitignore</h3><p>Add a .gitignore file in your root directory to exclude files from Git tracking.</p><p>📄 <strong>Example .gitignore:</strong></p><pre>node_modules/  <br>*.log  <br>.env  </pre><h3>🧬 Bonus: Rebase for Clean History</h3><p>To squash multiple commits before merging:</p><pre>git rebase -i &lt;commitId&gt;<br>git rebase -i HEAD~3  # Squash last 3 commits</pre><p>Then use:<br> ✅ pick to keep the commit<br> 🔀 squash to combine commits</p><h3>🔍 Visualize Git Like a Pro</h3><p>Try interactive Git simulations here:<br> 🧠 <a href="https://learngitbranching.js.org/">https://learngitbranching.js.org/</a></p><h3>📚 Suggested Readings</h3><ul><li><a href="https://git-scm.com/docs/gittutorial">Git Documentation</a></li><li><a href="https://www.w3schools.com/git/">W3Schools Git Tutorial</a></li><li><a href="https://www.geeksforgeeks.org/git-tutorial/">GeeksForGeeks Git Tutorial</a></li><li><a href="https://www.tutorialspoint.com/git/index.htm">TutorialsPoint Git</a></li></ul><h3>🎥 Recommended Courses</h3><ul><li><a href="https://youtu.be/RGOj5yH7evk">Git &amp; GitHub for Beginners Crash Course</a></li><li><a href="https://youtu.be/apGV9Kg7ics">Complete Git and GitHub for Beginners</a></li><li><a href="https://youtu.be/8JJ101D3knE">Git in 1 Hour</a></li></ul><h3>💡 Wrapping Up</h3><p>Git is a powerful tool that every developer should master. Start with the basics, practice regularly, and soon you’ll be managing projects like a pro! <strong>Got questions? Drop them in the comments!</strong></p><p><strong>Give a clap if you like it 👏.</strong></p><p>📌 <strong>Check out more of my content </strong><a href="https://medium.com/@padmesh2224"><strong>here</strong></a><strong> on Medium and connect with me:</strong><br> 🔗 <strong>LinkedIn</strong>: <a href="https://www.linkedin.com/in/padmesh-sharma">Padmesh Sharma</a><br> 📅 <strong>Book a 1:1 Chat</strong>: <a href="https://topmate.io/padmesh_sharma">topmate.io/padmesh_sharma</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=fbe502385f98" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[ First Corporate Experience as an SDE Intern? Here’s Your Ultimate Survival Guide!]]></title>
            <link>https://medium.com/@padmesh2224/first-corporate-experience-as-an-sde-intern-heres-your-ultimate-survival-guide-33ef04a5d97f?source=rss-64f4d52198ee------2</link>
            <guid isPermaLink="false">https://medium.com/p/33ef04a5d97f</guid>
            <category><![CDATA[sde-intern]]></category>
            <category><![CDATA[ppo]]></category>
            <category><![CDATA[corporate]]></category>
            <category><![CDATA[software-development]]></category>
            <dc:creator><![CDATA[Padmesh Sharma]]></dc:creator>
            <pubDate>Thu, 15 May 2025 09:38:46 GMT</pubDate>
            <atom:updated>2025-05-16T07:19:43.661Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*zMo_dQM4WnWJISo5mDMv4Q.png" /></figure><p>Stepping into your first corporate internship as an SDE (Software Development Engineer) is like entering a new game level-exciting, challenging, and packed with opportunities. Whether you’re already in or just gearing up, this blog will walk you through exactly what you need to thrive in your first SDE internship experience. ✨</p><p>Let’s dive into how you can <strong>prepare, perform, and potentially get that PPO (Pre-Placement Offer)!</strong></p><h3>🛠️ 1. Before You Step In: Prepare Like a Pro</h3><p>Think of this phase as warming up before a big match.</p><h3>✅ What You Must Brush Up:</h3><ul><li><strong>DSA (Data Structures and Algorithms): </strong>Practice on LeetCode, CodeStudio, or HackerRank. Even if your internship is project-based, strong problem-solving skills will help.</li><li><strong>Know Your Tech Stack: </strong>If you know the tools/languages the company uses (e.g., React, AWS, Python), get comfortable with them.</li><li><strong>Git &amp; Version Control: </strong>Learn basic Git commands (clone, commit, push, pull, merge).</li><li><strong>System Design Basics: </strong>Especially useful for 6-month internships.</li><li><strong>Project Understanding: </strong>If you’ve been given any hint about your project, research beforehand.</li></ul><h3>🧠 Soft Skills Prep:</h3><ul><li><strong>Communication: </strong>Practice writing concise emails and speaking confidently.</li><li><strong>Time Management: </strong>Use tools like Google Calendar or Notion to stay organized.</li></ul><h3>⏳ 2. Managing Work: From TODOs to Triumphs</h3><p>Welcome to the land of <strong>Jira boards, Slack pings, and unexpected bugs.</strong> 🎯</p><h3>🔍 Top Tips to Stay on Top:</h3><ul><li><strong>Use a Task Tracker</strong> : The Jira + Notion combo is golden. Track tasks in a personal Kanban board (To Do → Doing → Done).</li><li><strong>Follow Agile Basics: </strong>Understand what a sprint, story point, and stand-up are.</li><li><strong>Communicate Wisely</strong>: If you’re blocked, speak up early.</li></ul><h3>🤝 Meet Your Mentor:</h3><ul><li>Understand team goals.</li><li>Ask questions.</li><li>Don’t hesitate to say, “I’m not sure, but I’ll figure it out.”</li></ul><h3>🕑 2-Hour Rule:</h3><p>If stuck for more than 1–2 hours:</p><ul><li>Document what you tried.</li><li>Then ask for help.</li></ul><p>Example:</p><blockquote><em>“I tried X and Y to debug the API timeout, but the logs show Z. Any pointers?”</em></blockquote><h3>🧾 Stand-Up Script:</h3><ul><li><strong>Yesterday:</strong> I worked on [Task A].</li><li><strong>Today:</strong> I’ll tackle [Task B].</li><li><strong>Blockers:</strong> [Issue] investigating.</li></ul><h3>🔥 Pro Move:</h3><p>Set a <strong>weekly 15-minute recap</strong> with your mentor. Share:</p><ul><li>✔ Wins</li><li>❌ Challenges</li><li>🔮 Next Steps</li></ul><h3>🧍‍♀️🧍‍♂️ 3. Office Ethics: Because Culture &gt; Code</h3><p>Even the best coder can lose opportunities if they don’t play well with the team.</p><h3>🚦 The Unwritten Rules:</h3><ul><li>Be <strong>on time: </strong>Whether online or in-office.</li><li>Dress <strong>decently: </strong>Smart casuals are usually safe.</li><li>Be <strong>respectful: </strong>To your manager, peers, and even the pantry staff.</li><li><strong>No gossip. No politics.</strong> Ever.</li></ul><h3>📢 Communication Etiquette:</h3><ul><li><strong>Slack/Teams:</strong> Be polite, clear, and professional.</li><li><strong>Emails:</strong> Use proper greetings (e.g., <em>Hi [Name]</em> instead of <em>Hey</em>).</li><li><strong>Meetings:</strong> Mute when not speaking, avoid multitasking.</li></ul><h3>🚀 Networking:</h3><ul><li><strong>Build Relationships: </strong>Lunch with teammates, attend company events.</li><li><strong>Ask for Feedback: </strong>Shows you’re eager to improve.</li></ul><h3>💰 4. Money Talks: Start Investing Early</h3><p>Your first stipend might feel like a jackpot. But here’s the trick: <strong>don’t just spend , invest.</strong></p><h3>🔑 Smart Money Moves:</h3><ul><li><strong>Open a Demat account: </strong>Zerodha, Groww, or Upstox are beginner-friendly.</li><li><strong>Mutual Funds (SIP): </strong>Start small (₹500–₹1000/month).</li><li><strong>Avoid FOMO Trading: </strong>Crypto &amp; stocks can wait till you learn properly.</li><li><strong>Avoid EMIs: </strong>That shiny phone can wait.</li></ul><h3>🎓 Upskill Investments:</h3><p>Use part of your stipend for:</p><ul><li>Courses (Coursera, Udemy)</li><li>Certifications (AWS, GCP)</li></ul><h3>🏆 5. How to Bag that PPO (Pre-Placement Offer)</h3><p>This is the endgame for many interns. Here’s how to leave a mark:</p><h3>🚀 Go from Intern to Indispensable:</h3><ul><li><strong>Own a Project: </strong>Take initiative, don’t just follow instructions.</li><li><strong>Deliver Impact:</strong> Even if your task is small, do it exceptionally well.</li><li><strong>Ask Questions: </strong>But don’t ask Google-able stuff.</li><li><strong>Document Your Work: </strong>Maintain a daily log of contributions.</li><li><strong>Be Visible: </strong>Speak up in meetings (with value), share updates.</li><li><strong>Ask for Feedback: </strong>And implement it quickly.</li></ul><h3>👨‍💻 Be a Team Player:</h3><ul><li>Help colleagues when possible.</li><li>Volunteer for extra tasks (bug fixes, documentation).</li></ul><h3>📢 Showcase Your Impact:</h3><p>Before the final review, prepare a presentation:</p><ul><li>What you built.</li><li>Challenges faced &amp; how you solved them.</li><li>Business impact (e.g., improved performance by X%).</li></ul><h3>🔥 Bonus Tip:</h3><p>Schedule a 1:1 with your mentor or manager <strong>mid-internship</strong> and ask:</p><blockquote><em>“What can I do better to be PPO-worthy?”</em></blockquote><p>This shows initiative and a growth mindset.</p><h3>💡 <strong>Final Thoughts</strong></h3><p>You don’t need to have all the answers from day one. What matters most is your willingness to learn, adapt, and grow. This journey is yours make it count, one commit at a time. 💻✨</p><h3>💬 <strong>Have something on your mind?</strong></h3><p>Whether it’s a small doubt or a big worry, drop a comment . I’d love to hear your thoughts and help where I can. And hey, don’t forget to share this with someone stepping into their own SDE internship adventure.</p><p>If this guide gave you even one “aha!” moment, show it some love 👏.<br>Check out more of my content here on <a href="https://medium.com/@padmesh2224">Medium</a> and connect with me:</p><p>📌 <strong>LinkedIn</strong>: <a href="https://www.linkedin.com/in/padmesh-sharma">Padmesh Sharma</a><br> 📅 <strong>Book a 1:1 Chat</strong>: <a href="https://topmate.io/padmesh_sharma">topmate.io/padmesh_sharma</a></p><p>Wishing you growth, grit, and great code ahead! 🚀</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=33ef04a5d97f" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[ORACLE On-Campus Full-Time Interview Experience (SELECTED!)]]></title>
            <link>https://medium.com/@padmesh2224/oracle-on-campus-full-time-interview-experience-selected-0290a2356507?source=rss-64f4d52198ee------2</link>
            <guid isPermaLink="false">https://medium.com/p/0290a2356507</guid>
            <category><![CDATA[selected]]></category>
            <category><![CDATA[on-campus]]></category>
            <category><![CDATA[full-time]]></category>
            <category><![CDATA[oracle]]></category>
            <category><![CDATA[interview-experience]]></category>
            <dc:creator><![CDATA[Padmesh Sharma]]></dc:creator>
            <pubDate>Tue, 11 Mar 2025 05:59:30 GMT</pubDate>
            <atom:updated>2025-05-25T19:21:08.831Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*8Yo04Hu-1BB5Xdkb.jpg" /></figure><p>Hi everyone! 👋<br>I’m <strong>Padmesh Sharma</strong>, a final-year Computer Science student at <strong>Banasthali Vidyapith</strong>, and I’m excited to share my incredible journey of securing<strong> a Full-Time Role at Oracle</strong>!</p><p>Through this blog, I’ll walk you through every stage of the process, from the initial application to the final interview round. If you’re gearing up for Oracle’s recruitment or any similar process, I hope my experience helps you prepare better and boosts your confidence.</p><p>Let’s get started! 🚀</p><h3>📄 Application Process and Eligibility Criteria</h3><p>The Oracle campus recruitment drive kicked off on <strong>September 25, 2024</strong>, and was open to students graduating in <strong>2025</strong>. The company was hiring for both <strong>SDE Internship</strong> (6 months) and <strong>Full-Time roles</strong>, which made this drive particularly exciting for us.</p><p>Here are the eligibility criteria for this drive:</p><p><strong>Eligible Courses</strong>:</p><ul><li>B.Tech (CS, IT, ECE, EEE -Batch 2025)</li><li>M.Tech (CS, AI -Batch 2025)</li></ul><p><strong>CGPA Requirements</strong>:</p><ul><li>B.Tech: <strong>7.0 and above</strong></li><li>M.Tech: <strong>8.5 and above</strong></li></ul><p>A total of <strong>300+ students</strong> applied, making this a competitive process right from the start.</p><h3>🖥️ Round 1: Online Coding Assessment</h3><p>📅 <strong>Date</strong>: September 29, 2024<br> 📍 <strong>Platform</strong>: Hackerrank<br> ⏳ <strong>Duration</strong>: 100 minutes</p><p>The assessment tested our skills in <strong>Data Structures and Algorithms (DSA)</strong>, technical knowledge, and general aptitude. Here’s a breakdown of the sections:</p><ol><li><strong>Coding Problem</strong>:</li></ol><p>We were given one <strong>DSA question</strong> (medium to hard level) based on <strong>Dynamic Programming (DP)</strong>. The problem required finding the <strong>number of partitions of an array</strong> meeting certain conditions.</p><p>I managed to pass all test cases by <strong>optimized solution using DP</strong>.</p><p><strong>2. Technical MCQs</strong>:</p><p>This section had 20 multiple-choice questions covering topics like <strong>DBMS</strong>, <strong>Java</strong>, <strong>Python</strong>, and basic <strong>OOP concepts</strong>.</p><p><strong>3. Aptitude &amp; Communication</strong>:</p><p>15 questions on <strong>logical reasoning, basic mathematics</strong>, and problem-solving skills. This section felt manageable with regular practice.</p><blockquote>💡 <strong>Pro Tip</strong>: Practice <strong>medium-to-hard level DSA problems</strong> and revise your core subjects. Familiarity with platforms like Hackerrank or LeetCode is a huge plus.</blockquote><h3>🎤 Pre-Placement Talk (PPT)</h3><p>📅 <strong>Date</strong>: October 4, 2024</p><p>Before the interviews, Oracle organized an insightful <strong>Pre-Placement Talk</strong> to introduce us to the company’s culture, values, and expectations. The speakers included employees from diverse teams, and they emphasized Oracle’s commitment to innovation and employee growth.</p><p>The session inspired me to give my best in the upcoming rounds and reassured me that this was the right place for my career aspirations.</p><h3>💻 Round 2: Technical Interview I</h3><p>📅 <strong>Date</strong>: October 7, 2024<br> 📍 <strong>Mode</strong>: On-Campus (In-person)<br> ⏳ <strong>Duration</strong>: 1 hour</p><p>This was an in-person technical interview, and I began with a brief <strong>introduction</strong>, highlighting my academic background, internships, and technical projects.</p><p>The interviewer then tested my problem-solving skills by asking a <strong>stack-based DSA question</strong>. Here’s how it went:</p><ol><li>I started with the <strong>brute force approach</strong>, explaining my logic.</li><li>The interviewer encouraged me to <strong>optimize the solution</strong>. I shared an <strong>O(n) approach</strong> using auxiliary stacks, which worked well.</li><li>I coded the solution, and we discussed its <strong>time and space complexities</strong>.</li></ol><p>We also covered topics like <strong>DBMS joins, SQL queries</strong>, and <strong>System Programming concepts</strong>. Towards the end, the interviewer asked me to explain one of my projects. He seemed impressed by the practical challenges I solved during development.</p><h3>🧑‍💻 Round 3: Technical Interview II</h3><p>⏳ <strong>Duration</strong>: 1.5 hours</p><p>This round was more in-depth, focusing on advanced problem-solving and my ability to tackle <strong>real-world scenarios</strong>.</p><p>The interviewer asked me to solve <strong>three coding problems</strong>, each with increasing complexity:</p><ol><li><strong>Graphs</strong>: A shortest path problem.</li><li><strong>Arrays</strong>: An optimization challenge with subarrays.</li><li><strong>Linked Lists</strong>: A classic reversal problem with a twist.</li></ol><p>I managed to code and test two out of three problems successfully. For the third, I was asked to explain my approach in detail and discuss potential optimizations.</p><p>We also discussed my <strong>internship experiences at BNY and Amazon</strong>. The interviewer asked me to describe the challenges I faced, which allowed me to demonstrate my problem-solving mindset.</p><h3>🤝 Round 4: HR Interview</h3><p>⏳ <strong>Duration</strong>: 20 minutes</p><p>The HR round was conversational and felt more like a casual discussion. The HR manager asked questions like:</p><ul><li>“Why do you want to join Oracle?”</li><li>“What are your long-term career goals?”</li><li>“How do you handle pressure in high-stakes situations?”</li></ul><p>I shared my interest in <strong>Oracle’s cloud solutions</strong> and its role in shaping global technology. I also emphasized how my <strong>coding skills, leadership experience, and adaptability</strong> align with the company’s values.</p><blockquote>💡 <strong>Pro Tip</strong>: Be genuine and confident while answering HR questions. They want to know your personality, not just your technical expertise.</blockquote><h3>📅 Timeline of Events</h3><ul><li><strong>Recruitment Announcement</strong>: September 25, 2024</li><li><strong>Online Assessment</strong>: September 29, 2024</li><li><strong>Pre-Placement Talk</strong>: October 4, 2024</li><li><strong>Interviews</strong>: October 7, 2024</li><li><strong>Selection Mail</strong>: October 7, 2024</li></ul><h3>🎉 Final Verdict: SELECTED!</h3><p>Out of the 300+ applicants, only <strong>4 candidates</strong> were selected for the Full-Time role, and I’m overjoyed to be one of them!</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/0*L-cuBAMG7ZmwVWzU.png" /></figure><h3>🌟 My Key Takeaways:</h3><ol><li><strong>Prepare Core Concepts</strong>: Strong basics in <strong>DSA, DBMS, and OOP</strong> are critical.</li><li><strong>Communication is Key</strong>: Explain your approach clearly during problem-solving.</li><li><strong>Stay Positive</strong>: If one question doesn’t go well, don’t let it affect the rest of the interview.</li><li><strong>Consistency Matters</strong>: Regular practice on platforms like LeetCode, Hackerrank, Codechef helps immensely.</li></ol><p>If you’re preparing for Oracle or similar companies, I hope my story inspires you! Feel free to reach out if you have any questions, I’d be happy to help. 😊</p><p>If you found this helpful, feel free to clap<strong>👏</strong>, and check out my other blogs <a href="https://medium.com/@padmesh2224">here</a>.</p><p><strong>LinkedIn</strong>: <a href="https://linkedin.com/in/padmesh-sharma">Padmesh Sharma</a></p><p>Good luck with your journey! 🚀</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=0290a2356507" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Goldman Sachs Women’s Possibilities Summit Finale 2023–2024]]></title>
            <link>https://medium.com/@padmesh2224/goldman-sachs-womens-possibilities-summit-finale-2023-2024-ebe6f23e97b5?source=rss-64f4d52198ee------2</link>
            <guid isPermaLink="false">https://medium.com/p/ebe6f23e97b5</guid>
            <category><![CDATA[summit]]></category>
            <category><![CDATA[national-finalist]]></category>
            <category><![CDATA[goldman-sachs]]></category>
            <category><![CDATA[women-in-tech]]></category>
            <category><![CDATA[women-in-finance]]></category>
            <dc:creator><![CDATA[Padmesh Sharma]]></dc:creator>
            <pubDate>Tue, 19 Mar 2024 20:50:12 GMT</pubDate>
            <atom:updated>2024-03-19T21:01:35.762Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/720/1*AiO_dOID7VRLwh3NU6RVjA.jpeg" /></figure><p>As I sit down to pen down my thoughts, I find myself overwhelmed with gratitude and excitement. Just a few days ago, On 15th March 2024, I had the privilege of being a part of the Goldman Sachs Women’s Possibilities Summit Finale 2023–24 in Bengaluru. It was an experience that not only broadened my horizons but also reaffirmed my belief in the power of diversity and inclusion in today’s world.</p><p>The journey began when I registered for the summit and submitted my pitch on why careers and empowerment are crucial topics in today’s discourse. Little did I know that this simple act would lead me to an unforgettable adventure.</p><p>Arriving in Bengaluru, I was greeted with warmth and enthusiasm. Goldman Sachs had meticulously arranged every aspect of our trip, from travel to accommodation, ensuring our comfort and convenience throughout.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*xNNDFhv-NrePOOEWl_nU-w.jpeg" /></figure><p>The summit itself was a whirlwind of learning and inspiration. From insightful sessions with industry leaders like Shalini Satish Babu, and Divya Mathai to engaging presentations and mentoring sessions with Sandeep Jindal, and Reshma Joshi, every moment was a treasure trove of knowledge.</p><p>But beyond the sessions and presentations, what truly stood out was the sense of community and camaraderie among participants. We came from diverse backgrounds, representing various facets of society, yet we were united by a common goal — to drive positive change in the finance industry through diversity and inclusion.</p><p>As I reflect on my experience, I am filled with gratitude for the opportunities that have come my way. The summit has not only enriched my understanding of finance but has also connected me with a network of like-minded individuals who share my passion for breaking barriers and embracing diversity.</p><p>To Goldman Sachs and everyone involved in organizing this incredible event, I extend my heartfelt thanks. Your commitment to fostering diversity and inclusion is truly commendable, and I am honored to have been a part of it.</p><p>As I return home, I carry with me not just newfound knowledge and insights but also a sense of purpose and determination. The journey doesn’t end here — it’s just the beginning of a new chapter in my career, fueled by the lessons learned and connections made at the Women’s Possibilities Summit.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*TWakHiRjGYxOdjiJ-Hs99g.jpeg" /></figure><p>Here’s to embracing new opportunities, breaking barriers, and championing diversity and inclusion together. The future is bright, and I am excited to be a part of it.</p><p>Thank you for reading!</p><p>If you found this helpful, feel free to clap<strong>👏</strong>, and check out my other blogs <a href="https://medium.com/@padmesh2224">here</a>.</p><p>For any questions or doubts, don’t hesitate to comment or reach out to me on <a href="https://www.linkedin.com/in/padmesh-sharma/">LinkedIn</a>.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ebe6f23e97b5" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[AMAZON On-Campus Internship Interview Experience (SELECTED!)]]></title>
            <link>https://medium.com/@padmesh2224/amazon-on-campus-internship-interview-experience-selected-ef6ad321b747?source=rss-64f4d52198ee------2</link>
            <guid isPermaLink="false">https://medium.com/p/ef6ad321b747</guid>
            <category><![CDATA[sde-intern]]></category>
            <category><![CDATA[amazon]]></category>
            <category><![CDATA[selected]]></category>
            <category><![CDATA[6-month-internship]]></category>
            <dc:creator><![CDATA[Padmesh Sharma]]></dc:creator>
            <pubDate>Tue, 20 Feb 2024 06:46:05 GMT</pubDate>
            <atom:updated>2024-02-20T06:46:05.260Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/875/0*x0aEstZ7Pz_DhLfN.PNG" /></figure><p>Hi everyone, Padmesh Sharma here, a pre-final year CS student at Banasthali Vidyapith. I’m excited to share my journey to landing a 6-month SDE Internship at Amazon India! In this article, I will detail my entire interview experience.</p><h3>Application and Eligibility</h3><p>On Feb 5th,2023, I received an e-mail from my college regarding the<strong> 6-month project Internship drive (from July 2024 to December 2024) of AMAZON for Batch 2025</strong>. Over 270 students from my university participated in the online assessment.</p><p><strong>✅Eligibility Criteria</strong></p><ol><li><strong>Eligible Courses: B.Tech (CS / IT) VI Semester of Batch 2025</strong></li><li><strong>Eligibility Criteria: 65% &amp; Above separately in 10th Class, 12th Class &amp; in the current B.Tech course (aggregate till the IV sem)</strong></li></ol><h3>Pre-Placement Talk (PPT)</h3><p>We had a Pre-Placement Talk (PPT), an inspiring session where a panel of Amazon employees discussed the Amazon Brand and experiences and shared their tips for success on <strong>7th February 2024.</strong></p><h3>Round 1: Online Coding Assessment</h3><blockquote>Platform: Hackerrank<br>Sections:</blockquote><blockquote>1.<strong> Coding (70 minutes) </strong>comprising two Data Structure and Algorithm (DSA) questions.</blockquote><blockquote>2. <strong>Work Style Assessment (20 minutes) </strong>centered around Amazon’s Leadership Principles.</blockquote><p>One of the questions was based on greedy algo (medium level) and the other one was <strong>strings—based </strong>(medium level). I was able to pass all the test cases for one question and half for another.</p><p>Then, One had to answer around 50 questions in the workstyle assessment. In each question, there were two options and one had to move the slider closer to the option preferred. Each page had around 8 questions. One was not able to switch between pages. It was an easy round as the questions were supposed to be answered intuitively. I would suggest that one read about <a href="https://www.amazon.jobs/en/principles">Amazon&#39;s leadership principles</a> before attempting this round. This round is very important, as I have seen many people taking it lightly and getting rejected. Amazon places an equal importance on Behavioral and Technical assessment.</p><blockquote>I got the mail regarding the interview round after the week of my online assessment. 30 students were chosen for online interviews in total. The next round was scheduled the next day!</blockquote><h3>Round 2: Technical Interview</h3><blockquote>This one-hour interview on Amazon Chime involved two interviewers.</blockquote><p>It started with the three of us introducing ourselves. Do craft a concise introduction that encapsulates the key milestones of your professional journey summarizing the crucial aspects of your career. Then he jumped into DSA. He explained that he would be asking 2 DSA questions and I had to code both.</p><p>I can not disclose the exact questions.</p><blockquote><strong>Question 1</strong>: It was a question revolved around the <strong>application of stacks.</strong></blockquote><blockquote><strong>Question 2</strong>: The problem was based on <strong>trees </strong>and their traversals.</blockquote><p>I ensured that I collected all the anticipated requirements before commencing the coding process. In this round, it was important to have <strong>constant communication</strong> with the interviewer. He cross-questioned my approach many times. We deliberated on the advantages and disadvantages of different strategies, and he provided me with valuable hints. I began by elucidating the brute force solution and then progressed to the optimized approach, detailing the associated time and space complexities. Subsequently, he asked me to code the refined solution. He seemed satisfied with my code. In the end, he asked if I had any questions for him.</p><p>This round of interview lasted for almost 1 hr!!</p><h4>📍Some of the things to keep in mind during the interview process:</h4><blockquote>Discuss your approach clearly and think out loud.</blockquote><blockquote>Be open to hints and discuss solutions collaboratively.</blockquote><blockquote>Amazon’s Leadership Principles are crucial.</blockquote><blockquote>Aim for the most optimized solution while explaining complexity.</blockquote><blockquote>Discuss time and space complexity beforehand.</blockquote><blockquote>Prepare insightful questions to show your genuine interest.</blockquote><h3>⌛Timeline for the Entire Process</h3><ul><li><strong>Info about campus placement drive of AMAZON→ 5th February 2024</strong></li><li><strong>Online PPT Date &amp; Time: 7th February, 2024</strong></li><li><strong>Online Assessment on Hackerrank → 9th February 2024</strong></li><li><strong>Mail of shortlisted students for the interview process -&gt; 14th February 2024</strong></li><li><strong>Online Interview → 15th February 2024</strong></li><li><strong>Selection Mail -&gt; 16th February 2024</strong></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/757/1*2MQzTyTf6vn6U9xYF2rbcA.png" /></figure><h3><strong>Verdict — SELECTED! ✨🎉</strong></h3><p>Out of 30 students who were selected after the Online assessment, only 3 students<strong> were selected for a 6-month Summer Internship at Amazon </strong>including me. I was overjoyed!</p><p>That wraps up my experience. Best of luck to everyone with your upcoming interviews! Thank you for reading!</p><p>If you found this helpful, feel free to clap<strong>👏</strong>, and check out my other interview experience<a href="https://medium.com/@padmesh2224/greetings-readers-45429f3e2b52"> here</a>.</p><p>For any questions or doubts, don’t hesitate to comment or reach out to me on <a href="https://www.linkedin.com/in/padmesh-sharma/">LinkedIn</a>.</p><p>All the best! ❤️</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=ef6ad321b747" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[BNY Mellon On-Campus Internship Interview Experience (SELECTED!)]]></title>
            <link>https://medium.com/@padmesh2224/greetings-readers-45429f3e2b52?source=rss-64f4d52198ee------2</link>
            <guid isPermaLink="false">https://medium.com/p/45429f3e2b52</guid>
            <category><![CDATA[selected]]></category>
            <category><![CDATA[interview-experience]]></category>
            <category><![CDATA[summer-internships]]></category>
            <dc:creator><![CDATA[Padmesh Sharma]]></dc:creator>
            <pubDate>Sat, 11 Nov 2023 09:32:55 GMT</pubDate>
            <atom:updated>2025-03-13T05:36:54.009Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*1FLk5MTTJb25_LzJoxwC4g.jpeg" /></figure><p>Greetings readers!</p><p>I am Padmesh Sharma, a pre-final year student currently pursuing my B.Tech. in Computer Science from Banasthali Vidyapith, Jaipur. I shall be interning with BNY Mellon as a <strong>Software Developer Engineering (SDE) Intern</strong> in the summer of 2024.</p><p>BNY Mellon recently visited my university campus (Banasthali Vidyapith) for the first time to hire summer interns for Software developer roles. The selection procedure included an initial Online Assessment, succeeded by two interview rounds. The first interview focused on technical aspects, while the second encompassed a combination of technical and HR-related discussions.</p><p><strong><em>Note: All were eliminatory rounds.</em></strong></p><h3>About the Company</h3><p>BNY Mellon, or The Bank of New York Mellon Corporation, is a worldwide financial services firm offering a diverse array of financial and investment services to individuals, institutions, and corporations. BNY Mellon is the largest and oldest financial institution in the United States, with a history dating back to the 18th century. It plays a crucial role in the worldwide financial sector by providing a wide array of services to meet the varied requirements of its clientele, spanning from individual customers to large institutional investors.</p><p><strong>✅Eligibility Criteria</strong></p><ol><li><strong>Eligible Courses: B.Tech (CS / IT) V Semester of Batch 2025</strong></li><li><strong>Eligibility Criteria: 7.5 CGPA and above in the current B.Tech course only (aggregate till IV sem)</strong></li><li><strong>NO ACTIVE / CURRENT BACKLOGS</strong></li></ol><h3>Round 1: Online Assessment (Coding Test)</h3><blockquote><strong>Duration: 120 minutes</strong></blockquote><p>The coding round was hosted on the hackerrank platform and comprised four coding questions(1 easy, 2 medium, and 1 hard level).</p><p>Problems were from these topics:<br>1. Tree: Easy(15/15)<br>2. Graph: Medium(8/15)<br>3. Array: Medium(11/15) <br>4. Graph: Hard(7/15)</p><p>Out of the 400 students who participated in the Online Coding Test, only 24 successfully advanced to the next stage, which is the Interview Round.</p><blockquote><strong>✨Tips to ace the OA round:</strong></blockquote><blockquote><strong>Aim to solve questions quickly to increase your chances of advancing to the interview rounds.</strong></blockquote><blockquote><strong>Ensure a thorough understanding of the problem statement. If you encounter difficulty with a problem, it might be beneficial to explore other challenges before investing excessive time in one.</strong></blockquote><h3>Round 2: Technical Interview</h3><blockquote>Duration 70 minutes</blockquote><p>The interviewer introduced herself to me. She asked for my introduction as well. After a brief introduction and project discussion, we dove straight into the Object-Oriented Programming (OOPs) concept. I was supposed to explain the OOPS concepts by implementing polymorphism, inheritance, abstraction, virtual function, etc. She asked few more follow-ups related to OOPs (like the significance of polymorphism, real-world use of OOPS, etc ) and I was able to answer all the follow-ups correctly.</p><p>Then, I was given a coding question to solve where I had to discuss the approach and write a neat code. I quickly came up with the optimal solution. I wrote the code within a few minutes as I was able to solve the problem optimally using <strong>set, map, and hashing algorithms</strong>. Then we deliberated on the time and space complexities of the optimal solution. I even dry-run through a test case.</p><p>Additionally, she tasked me with devising the most suitable data structure for certain scenarios and outlining the strategy for storing the data within that structure. She moved on to ask about different traversal techniques of graphs and trees and their real-life examples. The first question was on the Dijkstra Algorithm and the second one I didn’t remember.</p><p>Then, she asked some situation-based and behavioral questions.</p><p>Subsequently, The discussion moved on to Database Management Systems (DBMS), where I was questioned about some concepts. Next, I was asked to answer two SQL queries, one of medium difficulty and another slightly tricky.</p><blockquote>After the lunch break, around 1 hour later I got to know that I was selected and my next round would be in 10 minutes!</blockquote><blockquote>After the first interview round, 14 students, including me, were shortlisted for the next interview stage.</blockquote><h3>Round 3: (Technical + HR Round)</h3><blockquote>Duration 30 minutes</blockquote><p>This round was taken by the director of BNY Mellon.</p><p>The interview started with a formal introduction of him followed by mine. Then we discussed the second project I am currently working on. Then, he asked some questions about my current project. He was trying to sense my thinking power by asking about more features I could add. I was unable to grasp it at first but after two guesses of different features and he gave me the hint, it struck me and I answered it.</p><p>We discussed my background and other details about his too. Then the interview ended when he finally asked me about my joining the office at Pune or Chennai as an upcoming summer intern.</p><blockquote><strong>✨ Important tips for the Interview Round</strong></blockquote><blockquote>Practice DSA daily.</blockquote><blockquote>Keep your introduction well-prepared beforehand.</blockquote><blockquote>Learn about various Time and Space complexities.</blockquote><blockquote>Start the DSA solution with a brute-force approach first. Then optimize your code.</blockquote><blockquote>Explain your approaches before the actual implementation.</blockquote><blockquote>Prepare your projects well.</blockquote><blockquote>Be prepared for follow-up questions.</blockquote><h3><strong>⌛Timeline for the Entire Process</strong></h3><ul><li><strong>Info about campus placement drive of BNY MELLON→ 5th October 2023</strong></li><li><strong>Deadline for submitting the Application → 6th October 2023</strong></li><li><strong>Online Assessment on Hackerrank → 10th October, 2023</strong></li><li><strong>Mail of shortlisted students for the interview process -&gt; 17th October 2023</strong></li><li><strong>In-person Interview → 18th October,2023</strong></li><li><strong>Selection Mail -&gt; 19th October,2023</strong></li></ul><figure><img alt="" src="https://cdn-images-1.medium.com/max/898/1*pdYNy_UmRk3OykRnY53wkQ.png" /></figure><p><strong>Verdict — SELECTED! ✨🎉</strong></p><p>Out of 14 students who were selected after 2nd interview Round, 13 students<strong> were selected for a 2-month Summer Internship at BNY Mellon </strong>including me.</p><p><strong>Consider Clapping 👏 if this article helped you :)</strong></p><p>If you have additional questions or any concerns, please don’t hesitate to reach out to me.</p><p><a href="https://www.linkedin.com/in/padmesh-sharma/">LinkedIn</a></p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=45429f3e2b52" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[TEXT CLASSIFICATION USING ML]]></title>
            <link>https://medium.com/@padmesh2224/text-classification-using-ml-17b756558d49?source=rss-64f4d52198ee------2</link>
            <guid isPermaLink="false">https://medium.com/p/17b756558d49</guid>
            <dc:creator><![CDATA[Padmesh Sharma]]></dc:creator>
            <pubDate>Fri, 22 Sep 2023 18:07:41 GMT</pubDate>
            <atom:updated>2023-11-11T11:52:27.643Z</atom:updated>
            <content:encoded><![CDATA[<figure><img alt="" src="https://cdn-images-1.medium.com/max/650/1*s6CLEpe_WtJH0bOaiflNWQ.png" /></figure><p>One of the most relevant and challenging problems in the domain of natural language processing is text classification or categorization, also popularly known as document classification. This task involves categorizing or classifying text documents into various (predefined) categories based on the inherent properties or attributes of each text document. This has applications in diverse domains and businesses, including email spam identification and news categorization.</p><p>The concept might seem simple and if you have a small number of documents, you can look at each document and gain some idea about what it is trying to indicate. Based on this knowledge, you can group similar documents into categories or classes. It starts getting more challenging once the number of text documents to be classified increases to several hundred thousand or millions. This is where techniques like feature extraction and supervised or unsupervised machine learning come in handy.</p><p>To formalize this problem more clearly, we will have a given set of classes and several text documents. Remember that documents are basically sentences or paragraphs of text. This forms a corpus. Our task is to determine which class or classes each document belongs to. Briefly, for a supervised classification problem, we need to have some labeled data that we can use for training a text classification model. This data is curated documents that are already assigned to some specific class or category beforehand. Using this, we essentially extract features and attributes from each document and make our model learn these attributes corresponding to each particular document and its class/category. This is done by feeding it to a supervised machine-learning algorithm. Of course, the data needs to be preprocessed and normalized before building the model. Once done, we follow the same process of normalization and feature extraction and then feed it to the model to predict the class or category for new documents. However, for an unsupervised classification problem, we do not have any labeled training documents and instead use techniques like clustering and document similarity measures to cluster documents based on their inherent properties and assign labels to them.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1005/1*lZvKgsqdlZRLLWJpMY5eKA.png" /></figure><h3><strong>TEXT CLASSIFICATION</strong></h3><p>Text or document classification is defined as the process of assigning text documents into one or more classes or categories, assuming that we have a predefined set of classes. Documents are textual documents and each document can contain a sentence or even a paragraph of words. A text classification system can successfully classify each document to its correct class(es) based on the inherent properties of the document.</p><p>Mathematically, we can define it as, given some description and attributes d for a document D, where d ∈ D, and given a set of predefined classes or categories, C = {c1, c2, c3, …, cn}. The actual document D can have many inherent properties and attributes, which lead it to be an entity in a high-dimensional space. Using a subset of that space with a limited set of descriptions and features depicted by d, we should be able to successfully assign the original document, D to its correct class Cx using a text classification system T. This can be represented by T : D → Cx. We can see there are several documents that can be assigned to various categories of food, mobile phones, and movies. Initially, these documents are all present together, just like a text corpus has various documents in it. Once it goes through a text classification system, which is represented as a black box, we can see that each document is assigned to one specific class or category we defined previously. The documents are just represented by their names in real data; they can contain descriptions of each product, specific attributes like movie genre, product specifications, constituents, and much more, which are properties that can be used as features in the text classification system, to make document identification and classification easier.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/874/1*TvHW9_ict98WKPc0JS-vHA.png" /></figure><h3>TYPES OF TEXT CLASSIFICATION</h3><p>There are various types of text classification. The two major types that are based on the type of content that makes up the documents. They are as follows.</p><p>● <strong>Content-based classification:</strong> Content-based classification is the type of text classification where priorities or weights are given to specific subjects or topics in the text content, which help determine the class of the document. A conceptual example is that of a book with more than 30% of the content about food preparation. It can be classified under cooking/recipes.</p><p>● <strong>Request-based classification: </strong>Request-based classification is influenced based on user requests and is targeted toward specific user groups and audiences. This type of classification is governed by specific policies and ideals based on user behavior and decisions. This type of classification is often used in applications like chatbots, virtual assistants, customer support systems, and information retrieval systems to understand and respond to user requests effectively.</p><p>There are several variants of text classification tasks, based on the number of classes to predict and the nature of predictions. These types of classification are based on the dataset, the number of classes/ categories pertaining to that dataset, and the number of classes that can be predicted on any data point. They are as follows:</p><p>● <strong>Binary classification: </strong>Binary classification is when the total number of distinct classes or categories is two and any prediction can contain either one of those classes.</p><p>● <strong>Multi-class classification:</strong> Multiclass classification is also known as multinomial classification and refers to a problem where the total number of classes is more than two and each prediction gives one class or category, which can belong to any of those classes.</p><p>● <strong>Multi-label classification:</strong> Multilabel classification refers to problems where each prediction can yield more than one outcome/predicted class for any data point.</p><h3><strong>AUTOMATED TEXT CLASSIFICATION</strong></h3><p>To automate text classification, we use several machine-learning techniques and concepts. There are two main types of machine learning techniques that are relevant to solving this problem. They are as follows:</p><p>●<strong> Unsupervised machine learning:</strong> Unsupervised learning refers to specific machine learning techniques or algorithms that do not require any prelabeled training data samples to build a model. The focus is more on pattern mining and finding latent substructures in the data rather than predictive analytics. We usually have a collection of data points, which could be textual or numeric depending on the problem we are trying to solve. We extract features from each of the data points using a process known as feature engineering and then we feed the feature set for each data point into our algorithm and try to extract meaningful patterns from the data, like trying to group together similar data points using techniques like clustering or summarizing documents based on topic models. This is extremely useful in text document categorization and is also called document clustering, where we cluster documents into groups based on their features, similarity, and attributes, without training any model on previously labeled data.</p><p>● <strong>Supervised machine learning:</strong> Supervised learning refers to specific machine learning techniques or algorithms that are trained on prelabeled data samples, known as training data, and corresponding training labels/classes. Features are extracted from this data using feature engineering and each data point has its own feature set and corresponding class/label. The algorithm learns various patterns for each type of class from the training data. Once this process is complete, we have a trained model. This model can then be used to predict the class for future test data samples once we feed their features to the model. Thus, the machine has actually learned, based on previous training data samples, how to predict the class for new unseen data samples. There are two major types of supervised learning algorithms, described as follows: Classification: Supervised learning algorithms are known as classification when the outcomes to be predicted are distinct categories, thus the outcome variable is a categorical variable in this case. Examples are news categories and movie genres Regression: Supervised learning algorithms are known as regression algorithms when the outcome we want to predict is a continuous numeric variable. Examples are house prices and weather temperature.</p><h3><strong>TEXT CLASSIFICATION SYSTEM</strong></h3><p>To build a text classification system, we need to make sure, we have our source of data and retrieve that data so that we can start feeding it to our system. The following main steps outline a typical workflow for a text classification system, assuming that we have our dataset already downloaded and ready to be used.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/744/1*g9Lut1wPcjEOi8m2tlfpzQ.png" /></figure><p>● <strong><em>Data Retrieval</em></strong> Obviously, the first step in any data science or machine learning pipeline is to access and retrieve the data necessary for our analysis and for building machine learning models.</p><p>● <strong><em>Data Preprocessing and Normalization</em></strong> Now we need to think about the general text preprocessing or wrangling stage. This involves cleaning, preprocessing, and normalizing text to bring text components like sentences, phrases, and words to some standard format. This enables standardization across our document corpus, which helps in building meaningful features and helps reduce noise, which can be introduced due to many factors like irrelevant symbols, special characters, XML and HTML tags, etc.</p><p>➢<em> Cleaning text:</em> Our text often contains unnecessary content, like HTML tags, which do not add much value when analyzing sentiment. Hence, we need to make sure we remove them before extracting features. The BeautifulSoup library does an excellent job of providing the necessary functions for this. Our strip_html_tags(…) function cleans and strips out HTML code.</p><p>➢ <em>Removing accented characters:</em> In our dataset, we are dealing with reviews in the English language so we need to make sure that characters with any other format, especially accented characters, are converted and standardized into ASCII characters. A simple example is converting é to e. Our remove_accented_chars(…) function helps in this respect.</p><p>➢ <em>Expanding contractions: </em>In the English language, contractions are basically shortened versions of words or syllables. Contractions pose a problem in text normalization because we have to deal with special characters like apostrophes and we also have to convert each contraction to its expanded, original form. Examples include do not to don’t and I would to I’d. The expand_ contractions(…) function uses regular expressions and various contractions mapped in the contractions.py module to expand all contractions in our text corpus.</p><p>➢ <em>Removing special characters: </em>Another important task in text cleaning and normalization is to remove special characters and symbols that often add to the extra noise in unstructured text. Our function remove_special_ characters(…) helps remove special characters.</p><p>➢ <em>Stemming or lemmatization: </em>Word stems are usually the base form of possible words that can be created by attaching affixes, like prefixes and suffixes, to the stem to create new words. This is known as inflection. The reverse process of obtaining the base form of a word is known as stemming. A simple example is “watches,” “watching,” and “watched,” which have the word root stem “watch” as the base form. The NLTK package offers a wide range of stemmers like the PorterStemmer and LancasterStemmer. Lemmatization is very similar to stemming, where we remove word affixes to get to the base form of a word. However, the base form in this case is known as the root word but not the root stem. The difference is that the root word is always a lexicographically correct word (present in the dictionary) but the root stem may not be correct. The function lemmatize_text(…) helps us in that respect.</p><p>➢<em> Removing stopwords:</em> Words that have little or no significance, especially when constructing meaningful features from text, are known as stopwords. These are usually words that end up having the maximum frequency if you do a simple term or word frequency in a document corpus. Words like “a,” “an,” “the,” and so on are stopwords. There is no universal stopword list but we use a standard English language stopwords list from NLTK. You can add your own domain-specific stopwords if needed. The function remove_stopwords(…) helps remove stopwords and retain words having the most significance and context in a corpus. We use all these components and tie them together in the function called normalize_corpus(…), which can be used to take a document corpus as input and return the same corpus with cleaned and normalized text documents.</p><p>● <strong><em>Building Train and Test Datasets</em></strong><em>:</em> To build a machine learning system, we need to build our models on training data and then test and evaluate their performance on test data. Hence, we split our dataset into train and test datasets. We take a training dataset: test dataset split of 67%/33% of the total data.</p><p>● <strong>Feature Engineering Technique: </strong>In machine learning terminology, features are unique measurable attributes or properties for each observation or data point in a dataset. Features are usually numeric in nature and can be absolute numeric values or categorical features that can be encoded as binary features for each category in the list using a process called one-hot encoding. They can be represented as distinct numerical entities using a process called label-encoding. The process of extracting and selecting features is both an art and a science and this process is called feature extraction or feature engineering. Feature engineering is very important and is often known as the secret sauce to creating superior and better-performing machine learning models. Extracted features are fed into machine learning algorithms for learning patterns that can be applied to future new data points for getting insights. These algorithms usually expect features in the form of numeric vectors because each algorithm is at heart a mathematical operation of optimization and minimizing loss and error when it tries to learn patterns from data points and observations. Hence, with textual data comes the added challenge of figuring out how to transform and extract numeric features from textual data.</p><p>➢<em> </em><strong><em>Traditional Feature Engineering Models</em></strong> Traditional feature engineering is the process of creating new input features for a machine learning model by transforming, combining, or extracting information from the raw data.</p><p>❖ <em>Bag of Words (term frequency) model</em>: The Bag of Words model represents each text document as a numeric vector where each dimension is a specific word from the corpus and the value could be its frequency in the document, occurrence (denoted by 1 or 0), or even weighted values. The model’s name is such because each document is represented literally as a bag of its own words, disregarding word orders, sequences, and grammar.</p><p>❖<em> Bag of N-Grams model:</em> An N-gram is basically a collection of word tokens from a text document such that these tokens are contiguous and occur in a sequence. Bi-grams indicate n-grams of order 2 (two words), tri-grams indicate n-grams of order 3 (three words), and so on. The Bag of N-Grams model is just an extension of the Bag of Words model so we can also leverage N-gram-based features.</p><p>❖ <em>TF-IDF model:</em> TF-IDF stands for Term Frequency-Inverse Document Frequency and it’s a combination of two metrics, term frequency (TF) and inverse document frequency (IDF). This technique was originally developed as a metric for showing search engine results based on user queries and has become part of information retrieval and text feature extraction</p><p>➢<strong> <em>Advanced Feature Engineering Models</em></strong> Advanced feature engineering is employed in machine learning when traditional feature engineering techniques may not be sufficient to capture complex patterns or relationships in the data. While traditional feature engineering focuses on creating basic features from raw data, advanced feature engineering seeks to extract more nuanced, domain-specific, or high-level features that can lead to improved model performance or better data understanding.</p><p>❖ <em>Word2Vec model:</em> This model was created by Google in 2013 and is a predictive deep learning-based model to compute and generate high-quality, distributed, and continuous dense vector representations of words that capture contextual and semantic similarity.</p><p>❖ <em>GloVe model:</em> The GloVe model stands for Global Vectors. It’s an unsupervised learning model that can be used to obtain dense word vectors, similar to Word2Vec. However, the technique is different and training is performed on an aggregated global word-word cooccurrence matrix, giving us a vector space with meaningful substructures.</p><p>❖<em> FastText model:</em> The FastText model was introduced by Facebook in 2016 as an extension and an improvement of the vanilla Word2Vec model. FastText is a framework for learning word representations and performing robust, fast, and accurate text classifications. The Word2Vec model typically ignores the morphological structure of each word and considers a word a single entity. The FastText model considers each word a Bag of Character n-grams. 9</p><p>● <strong><em>Classification Models:</em></strong> Classification models are supervised machine learning algorithms that are used to classify, categorize, or label data points based on what it has observed in the past. Each classification algorithm is a supervised learning algorithm so it requires training data. This training data consists of a set of training observations where each observation is a pair that consists of an input data point, which is usually a feature vector like we observed earlier, and a corresponding output outcome for that input observation. There are three stages that classification algorithms go through during the modeling phase:</p><p>● <strong><em>Training:</em></strong> Training is the process where the supervised learning algorithm tries to infer patterns out of the training data so that it can identify which patterns lead to a specific outcome. These outcomes are often known as the class labels/class variables/response variables. We usually carry out the process of feature extraction or feature engineering to derive meaningful features from the raw data before training. These feature sets are fed to an algorithm of our choice, which then tries to identify patterns from these features and their corresponding outcomes. The result of this is an inferred function known as a model or a classification model. This model is expected to be generalized enough from learning patterns in the training set so that it can predict the classes or outcomes for new data points in the future.</p><p>● <strong><em>Evaluation:</em></strong> Evaluation involves trying to test the prediction performance of our model to see how well it has trained and learned on the training dataset. For this, we use a validation dataset and test the performance of our model by predicting on that dataset and testing our predictions against the actual class labels, also called the ground truth. We also often use cross-validation where the data is divided into folds and a chunk of it is used for training and the remaining is used to validate the trained model. A point to remember is that we also tune the model based on the validation results to get to an optimal configuration, which yields the maximum accuracy and minimum error.</p><p>●<strong><em> Tuning:</em></strong> Tuning is also known as hyperparameter tuning or optimization, where we focus on trying to optimize a model to maximize its prediction power and reduce errors. Each model is at heart a mathematical function, which has several parameters determining model complexity, learning capability, and so on. These are known as hyperparameters because they cannot be learned directly from data and must be set prior to running and training the model. Hence, the process of choosing an optimal set of model hyperparameters such that the performance of the model yields good prediction accuracy is known as model tuning.</p><h3>CLASSIFICATION ALGORITHMS</h3><p>Typically, there are various classification algorithms, that we use shortly when building our classification models.</p><p>● Multinomial Naïve Bayes</p><p>● Logistic regression</p><p>● Support vector machines</p><p>● Random forest</p><p>● Gradient boosting machine</p><p>We will discuss more about the Logistic Regression classification algorithm further.</p><h3>LOGISTIC REGRESSION</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/401/1*slQzPJ6myX8geaBuRDTchQ.png" /></figure><p>The logistic regression model is actually a statistical model developed by statistician David Cox in 1958. It is also known as the logit or logistic model since it uses the logistic (popularly also known as sigmoid) mathematical function to estimate the parameter values. These are the coefficients of all our features such that the overall loss is minimized when predicting the outcome. However, we don’t focus on errors but more on maximizing the likelihood of the predicted values to the observed values using Maximum-Likelihood Estimation (MLE). Considering a binary classification problem of predicting two classes, a 0 or a 1, in the logistic model, the log odds (the logarithm of the odds) for the class/category labeled as 1 are basically the equation of the linear regression model (linear combination of one or more independent features, which can be categorical or continuous). However, we need to predict discrete classes or categories. Thus, the corresponding probability of the class labeled 1 can vary between 0 and 1, depicting the confidence of the prediction. The function that helps us convert the log odds to probability is the logistic function. The standard sigmoid or logistic function can be depicted mathematically by this formula:</p><p>Where e is the exponent (Euler’s number) and x indicates the typical equation, which can be derived from the linear regression equation where we try to estimate the coefficients of our features. This function typically looks like an S-shaped curve, as depicted — The standard unit of measurement for the log-odds scale is called a logit, from the logistic unit, hence we have the alternative names for this model. Consider a standard multiple linear regression model, depicted as follows: Such that {x1, x2, …, xn} are our features and we are trying to estimate the coefficients, {β1,β2, …,βn}. Considering we need to predict the categorical classes, we can represent this as the log odds, as follows using the logit of the probability p This means that if p is the probability of predicting a specific class, the odds of that is p/(1- p), which is basically the ratio of the favorable outcomes to the unfavorable outcomes. Likewise, the logit of p is basically the log-odds. Thus, we can mathematically derive this as follows If we want to get to the class probability values that the logistic regression model outputs for us, we can derive the following equation, which is the heart of the logistic regression model:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/226/1*9JMH7OVbOeNGF9QveQtK7A.png" /></figure><p>Finally, we can use MLE to optimize and estimate the optimal coefficients for each feature, which helps in maximizing the likelihood function. In Scikit-Learn, the LogisticRegression model can be leveraged to use the logistic regression model for classification. The solvers implemented in the LogisticRegression class are “liblinear”, “newton-cg”, “lbfgs”, “sag”, and “saga”. Each of them has its own distinct implementations. In the case of multi-class classification, just like in our problem, the training algorithm uses the one-vs-rest (OvR) scheme if the multi_class option is set to ovr and uses the cross-entropy loss if the multi_class option is set to multinomial.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/900/1*RWeSa7jW94YvIB3Q1hUOMw.png" /></figure><blockquote>Advantages of Logistic regression</blockquote><blockquote>● Interpretability</blockquote><blockquote>● Efficiency</blockquote><blockquote>● Low Variance</blockquote><blockquote>● Works well with Few Features</blockquote><blockquote>● Good for Probabilistic Predictions</blockquote><blockquote>Disadvantages of using Logistic Regression</blockquote><blockquote>● Linear Decision Boundary</blockquote><blockquote>● Limited Expressiveness</blockquote><blockquote>● Assumption of Independence</blockquote><blockquote>● Not Suitable for High-Dimensional Data</blockquote><blockquote>● No Natural Handling of Missing Data</blockquote><h3>APPLICATIONS OF TEXT CLASSIFICATION:</h3><p>Text classification and categorization are used in several real-world scenarios and applications. Some of them are as follows:</p><p>● <em>News Categorization:</em> Text classification is used to categorize news articles into topics such as politics, sports, technology, entertainment, and more. It helps news agencies and readers quickly locate articles of interest.</p><p>● <em>Spam Filtering: </em>Email services and messaging platforms employ text classification to identify and filter out spam or unwanted messages. Machine learning models can learn to recognize patterns in the content or metadata of messages associated with spam.</p><p>● <em>Music or Movie Genre Categorization: </em>In the entertainment industry, text classification can be used to automatically categorize music tracks or movies into genres. This helps streaming platforms recommend similar content to users.</p><p>● <em>Sentiment Analysis: </em>Sentiment analysis involves determining the sentiment or emotional tone of a piece of text, such as a customer review or social media post. It is widely used in market research, customer feedback analysis, and brand monitoring.</p><p>● <em>Language Detection:</em> Text classification is used to automatically detect the language of a given text. This is important for multilingual applications, such as web content localization, language-specific search engines, and translation services.</p><p>● <em>Customer Support and Chatbots: </em>Text classification is used to route customer support queries to the appropriate department or support agent based on the content of the message. Chatbots also use text classification to understand user queries and provide relevant responses.</p><h3>BENEFITS OF TEXT CLASSIFICATION</h3><p><strong><em>Automated Categorization:</em></strong> Text classification allows for the automated categorization of large volumes of text, saving time and resources compared to manual tagging and sorting.</p><p><strong><em>Improved Information Retrieval:</em></strong> It enhances information retrieval by enabling more precise search results, ensuring users can find relevant content quickly.</p><p><strong><em>Efficient Spam Filtering: </em></strong>It helps filter out spam emails and messages, reducing the clutter in inboxes and improving overall communication efficiency.</p><p><strong><em>Sentiment Analysis:</em></strong> Text classification aids in understanding public sentiment, which is valuable for businesses to gauge customer opinions and make data-driven decisions.</p><p><strong><em>Automation of Customer Support: </em></strong>Automated text classification can route customer inquiries to the appropriate support channels, improving response times and customer satisfaction.</p><p><strong><em>Fraud Detection: </em></strong>It assists in detecting fraudulent activities by analyzing text data associated with transactions, helping financial institutions and e-commerce platforms prevent fraud.</p><h3>LIMITATIONS OF TEXT CLASSIFICATION</h3><p><strong><em>Ambiguity:</em></strong> Natural language is often ambiguous, making it challenging to categorize text accurately, especially when the context is crucial.</p><p><strong><em>Training Data Requirements:</em></strong> Developing accurate text classifiers often requires substantial amounts of labeled training data, which can be time-consuming and expensive to acquire.</p><p><strong><em>Privacy Concerns:</em></strong> Analyzing and categorizing text data may raise privacy concerns, especially when dealing with sensitive personal information.</p><p><strong><em>Lack of Context:</em></strong> Text classifiers may struggle with understanding context and sarcasm, leading to incorrect categorization or sentiment analysis in certain cases.</p><p><strong><em>Multilingual Challenges:</em></strong> Handling multiple languages and dialects can be complex and require additional resources and expertise.</p><p><strong><em>Ever-Changing Data:</em></strong> Text classification models may become outdated as language and context evolve over time, requiring ongoing maintenance and updates.</p><p><strong><em>Complexity and Interpretability:</em></strong> Deep learning models used in text classification can be complex and difficult to interpret, making it challenging to understand how they arrive at their decisions.</p><h3>FUTURE OF TEXT CLASSIFICATION</h3><p>★<strong>Advanced Pre-trained Language Models: </strong>The proliferation of advanced pre-trained language models like GPT-4 will lead to more accurate and context-aware text classification, benefiting various applications, including sentiment analysis and content categorization.</p><p>★<strong>Multimodal Data Integration: </strong>Text classification will evolve to integrate text with images, audio, and video data, enabling more comprehensive and context-rich content understanding, especially in applications like social media analysis.</p><p>★<strong>Transfer Learning:</strong> The use of transfer learning, where models pre-trained on large datasets are fine-tuned for specific tasks, will reduce the need for extensive labeled data, accelerating the development of domain-specific solutions in text classification.</p><p>★<strong>Interdisciplinary Applications:</strong> Text classification will extend its reach into interdisciplinary domains, finding applications in healthcare, law, finance, and more. It will be used for tasks such as patient record analysis, e-discovery, contract analysis, fraud detection, and sentiment analysis.</p><h3><strong>CONCLUSION</strong></h3><p>Text classification using machine learning is a versatile and indispensable tool across a wide range of domains. It enables automated categorization of text documents, streamlining processes, and facilitating informed decision-making. In the realm of algorithms, choices abound, from traditional methods like Naive Bayes and Logistic Regression to cutting-edge deep learning models like Convolutional Neural Networks (CNNs) and Recurrent Neural Networks (RNNs). The selection depends on the specific nature of the text data and the problem at hand. Effective preprocessing, data cleaning, and model tuning are essential to achieve optimal performance. In recent years, significant strides in natural language processing (NLP), driven by pre-trained language models such as BERT and GPT, have raised the bar in text classification. In essence, text classification, bolstered by the advances in NLP and machine learning, continues to be a cornerstone of AI-driven applications. Its ability to categorize and understand text data empowers organizations to extract insights, improve user experiences, and automate tasks, making it an indispensable tool for the future.</p><blockquote>REFERENCES</blockquote><blockquote>➢Dipanjan Sarkar “Text Analytics with Python” A Practitioner’s Guide to Natural Language Processing Second Edition ➢https://www.researchgate.net/publication/228084521_Text_Classification_Using_Ma chine_Learning_Techniques ➢https://monkeylearn.com/blog/text-classification-machine-learning ➢https://www.datacamp.com/tutorial/text-classification-python ➢https://levity.ai/blog/text-classifiers-in-machine-learning-a-practical-guide</blockquote><blockquote>➢https://www.sciencedirect.com/topics/medicine-and-dentistry/logistic-regression-analysis 19</blockquote><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=17b756558d49" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>