<?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 Oluwabamigbe Oni on Medium]]></title>
        <description><![CDATA[Stories by Oluwabamigbe Oni on Medium]]></description>
        <link>https://medium.com/@BamiTega?source=rss-a6b2b33db1ec------2</link>
        <image>
            <url>https://cdn-images-1.medium.com/fit/c/150/150/2*1YM-PYNU8pEyv4wX6JRC-Q.jpeg</url>
            <title>Stories by Oluwabamigbe Oni on Medium</title>
            <link>https://medium.com/@BamiTega?source=rss-a6b2b33db1ec------2</link>
        </image>
        <generator>Medium</generator>
        <lastBuildDate>Thu, 19 Nov 2026 04:54:13 GMT</lastBuildDate>
        <atom:link href="https://medium.com/@BamiTega/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[Retrieving Evidence for Fake News Detection using Deep learning on AWS — Brief]]></title>
            <link>https://medium.com/@BamiTega/retrieving-evidence-for-fake-news-detection-using-deep-learning-on-aws-brief-84d63d790c61?source=rss-a6b2b33db1ec------2</link>
            <guid isPermaLink="false">https://medium.com/p/84d63d790c61</guid>
            <category><![CDATA[data-science]]></category>
            <category><![CDATA[aws]]></category>
            <category><![CDATA[nlp]]></category>
            <category><![CDATA[deep-learning]]></category>
            <category><![CDATA[fake-news]]></category>
            <dc:creator><![CDATA[Oluwabamigbe Oni]]></dc:creator>
            <pubDate>Thu, 26 Mar 2020 12:49:03 GMT</pubDate>
            <atom:updated>2020-03-27T10:03:26.643Z</atom:updated>
            <content:encoded><![CDATA[<h3>Retrieving Evidence for Fake News Detection using Deep learning on AWS — Brief</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*zmXyuWE8E9BJLCOIhUg1pg.png" /></figure><p><em>This article will set the scene for a series of articles on </em><strong><em>Retrieving evidence for Fake News detection using Deep learning</em></strong><em>.</em></p><p><em>The series will look at: how to detect fake news in an automated manner; which data science frameworks and tools to consider; how to work with large data sets — including recommendations; and precautions when using cloud infrastructure such as AWS.</em></p><p>I’ve always been fascinated with rumour’s impact on a group of people. At its most basic, every rumour is just a subject of conversation, or folk tale, but at its most severe, rumour can cause major societal impacts and dynamic shifts in favour of its benefactors.</p><blockquote>“In a lawsuit the first to speak seems right, until someone comes forward and cross-examines.” — Book of Proverbs</blockquote><p>Within this frame of thought, every rumour seems true until it is supported or negated. The type of rumour called Fake News is false news. It is also rumour that is specifically news. Six main concepts exist relating to fake news:</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*_ARCaylsHM-HC-PIr-8nnQ.png" /><figcaption>A comparison between concepts related to fake news from http://arxiv.org/abs/1812.00315</figcaption></figure><p>Each one with its own peculiarity in the authenticity, intention and news categories, as well as in its methods of detection.</p><p>During the 2016 US presidential elections Donald Trump drove the surge in popularity of the term “Fake News”. Also, in his first press conference as president, he called a CNN reporter “Fake News”. This caused:</p><ol><li>A dynamic shift in favour of Donald Trump.</li><li>A surge of fake news stories on the web.</li><li>An increased interest in automated processes that detect Fake News e.g. PolitiFact.</li></ol><p>Automated ways to detect Fake News can be divided into 4 main detection strategies. Each strategy is more proficient in finding certain types of rumour:</p><ol><li>Source-based Detection: analysis account roles in rumour dissemination around a subject.</li><li>Style-based Detection: identifies rumour by analysing its written format. This is good for finding disinformation.</li><li>Propagation-based Detection: identifies rumour by analysing its spread.</li><li>Content-based Detection: involves analysis types that focus on false knowledge in Fake News.</li></ol><p>In this study, we focus on misinformation and we use fact-checking with evidence retrieval to classify for misinformation. Fact-checking is a content-based detection strategy.</p><p>The type of data source is vital for a robust model that efficiently detects misinformation. There is a wide range of freely available data-sets to consider when taking up this problem. In addition to the <a href="http://fever.ai/">FEVER Shared Task</a> used in this study, some examples are the FakeNewsChallenge (FNC) data-set and the FootballTransferRumours (FTR-18) data-set.</p><h3>The FEVER Shared Task</h3><p>In the shared task, 50 annotators put together 185k labelled claims in a 5-way agreement, providing evidences for verifiable claims, with a substantial kappa agreement score of 69%.</p><p>For the task, every verdict that is given on a claim must be verified by the evidences extracted for that claim. A claim can be classified as <em>SUPPORTED, REFUTED </em>or<em> NOT ENOUGH INFO</em>. Evidences are extracted from Wiki articles. This is seen as the source of truth.</p><p>Below are the problem and solution formulation details of the Shared task:</p><figure><img alt="Example of the FEVER shared task applied to a claim." src="https://cdn-images-1.medium.com/max/1024/1*lXEtiP_emeB41JKSKIy53A.png" /><figcaption>Problem formulation of FEVER Shared task</figcaption></figure><figure><img alt="The FEVER Shared task in practice" src="https://cdn-images-1.medium.com/max/1024/1*ymPc-e--G1RUmEZT0vk2EA.png" /><figcaption>Solution formulation of FEVER Shared task</figcaption></figure><p>Our data-set sizes are:</p><ul><li>5M+ Introductory sections of Wiki pages — 7GB</li><li>185k+ claims — 32mb</li></ul><p>The files format are json lines. This saves space when stored on disc and can be read straight into a pandas dataframe for analysis. Due to the multiplicative data size during calculations across every claim per Wiki page, we need extra fire power. For this we call in our Big guns in the AWS suite to perform these calculations at scale.</p><p>We followed the standard 3 stage process proposed in the FEVER Shared task to solve the problem.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*2yb-Wh9B2AXWCQKnU0jK4Q.png" /><figcaption>Our approach to Fact-checking of the FEVER shared task</figcaption></figure><h4>Stage 1: Document Retrieval</h4><p>The document retrieval step involves extracting the top N most relevant documents to a claim from a corpus. it is important that document retrieval has a high recall allowing documents containing evidences required to be extracted early in the process, so evidences get a chance to filter through the process.</p><p>To improve recall, the Named Entity Recognition (NER) module provided in the python Spacy package is used to acquire named entities in each text. We then filter the documents and perform other transformations and data enrichment before applying TF-IDF to each wiki pages per claim, ranking documents according to importance, and retrieving the top N documents.</p><h4>Stage 2: Sentence Selection</h4><p>In Sentence Selection, we extract the top L most relevant sentences to a claim from the top N documents. The top N documents are split into sentences. Unlike Document Retrieval, Sentence Selection involves evidence retrieval for a claim. Hence, we would like to extract sentences that are highly interrelated to a given claim, because this far extends word importance provided by TF-IDF we use Word2Vec embeddings for word vectorization, we fed these into the LSTM variant architectures, classify and then rank the scores for our claim-sentence pairs, and then selecting the top L sentences.</p><h4>Stage 3: Claim Verification</h4><p>In Claim Verification, we classify each claim as <em>SUPPORTED, REFUTED </em>or<em> NOT ENOUGH INFO</em> using the evidences retrieved. Here each claim is concatenated with the top L sentences before it is fed either into an LSTM variant architecture, or a Machine Learning classifier(e.g. Random Forest). The ML classifier requires that morphological, grammatical and lexical features are extracted from the text. The chosen classifier provides the verdict.</p><p>Inline with the Shared task, the verdict provided is only correct if the required evidence for verification is present.</p><h3>Working with Large Datasets on AWS</h3><p>AWS services like SageMaker and Redshift were used to make the data processing and solution formations in these pipelines feasible. I will highlight how and where these services were used in following articles. For now, I will leave you with some learning, notes and takeaways from working with large data sets like the one encountered in this study.</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/1024/1*D4wpk92rXnHQ7LbX_jmRKw.png" /><figcaption>Notes for working with large data sets</figcaption></figure><p>Hope you liked the read. Please drop your comments below.</p><p>Keep well, till the next article!</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=84d63d790c61" width="1" height="1" alt="">]]></content:encoded>
        </item>
        <item>
            <title><![CDATA[Not Waterfall, but Lean and Agile in DevOps]]></title>
            <link>https://medium.com/@BamiTega/not-waterfall-but-lean-and-agile-in-devops-981ae483968a?source=rss-a6b2b33db1ec------2</link>
            <guid isPermaLink="false">https://medium.com/p/981ae483968a</guid>
            <category><![CDATA[agile]]></category>
            <category><![CDATA[devops]]></category>
            <category><![CDATA[lean]]></category>
            <dc:creator><![CDATA[Oluwabamigbe Oni]]></dc:creator>
            <pubDate>Thu, 05 May 2016 21:43:46 GMT</pubDate>
            <atom:updated>2016-05-05T21:51:35.879Z</atom:updated>
            <content:encoded><![CDATA[<p>Not Waterfall, but Lean and Agile in DevOps</p><h3>Introduction</h3><p>Everybody enjoys using software because it makes our lives and endeavours easier and more efficient. The need for software, although a narrowly debated topic, is under less scrutiny than the efficiency of software development processes itself. This is with respect to modern day software information systems. Software development processes are a set of intellectual and creative processes, based on complex decision-making that can affect the value and quality of developed software (I Sommerville, 2012). Due to the great need for software in today’s society, software process models have been developed to make software development for similar systems quicker and easier.</p><p>In 1970, the most widely known software development model was proposed although not named at birth, it was eventually called the waterfall model (P A. Laplante, 2004). This model involves five principal stages of software development used to produce a desired product. Its weakness however includes its rigid structure and its failure to respond to changing requirements from its customers. One truth behind its failure is that changing requirements is not an error but an integral part of building a system to a customer’s requirement. The fact that this was obviously not expected brought about the need for different approaches to software development that could succeed when the classic waterfall model failed.</p><p>There are different software development methodologies but the waterfall model was one of the methodology to be first adopted. New approaches to software development is not just a development model, but an entire framework and a methodology that involves people, processes and system development. All this came together to create shared values and principles that derives the culture. Other methodologies such as the lean, the agile and DevOps are only a few of the many different types of methodologies now available but have manage to cause a whole paradigm in the analysis, reasoning and development of software systems. The shared principles and differences between these frameworks allow these software development processes to stand out and provide a success rate that supersedes that of others.</p><h3>The waterfall model</h3><p>The waterfall model which is generally referred to as the classical waterfall model was first presented by Winston Royce in 1970 (Phillip et al., 2016). Initially standard software development models were seen as isolated processes. In his paper Winston proposed that software development model phases are intertwined and iterative activities, and that their iterations are not successive.</p><p>(I Sommerville, 2012)</p><p>The waterfall model is a plan-based model that involves five key stages, and are the software products life-cycle, each stage produces an artefact (a type of document) that is signed-off before the following stage is begun. The artefacts produced in the previous stages are also used in upcoming processes to create another artefact, the process is continued until the software is delivered, and maintenance occurs. Below is a breakdown of the waterfall process (I Sommerville, 2012).</p><figure><img alt="" src="https://cdn-images-1.medium.com/max/496/1*9r9r1ol3sz8GiAARXbmtVw.png" /><figcaption>(I Sommerville, 2012)</figcaption></figure><p>1. <strong>Requirements Definition:</strong> The requirement definition is the first stage of the development process. In this stage, a consultant from the development company consults with the customers and then derives the systems functions, constraints and characteristics and this is then broken down in a document called the system specification.</p><p>2. <strong>System and software design:</strong> The system specification is then used to model the customer requirements as a software system. Every single functional requirement is then attributed to a component of the system, allowing the overall system architecture of the system to be in sync with the customer’s requirements.</p><p><strong>3. Implementation and Unit testing: </strong>In this stage, code is written for each component according to the specified design or at least there is an attempt to write code according to the specified design. As we know with everything in reality, the perfect scenario rarely occurs, which forces the need to be able to ideally deal with this change in the best possible way. Each unit that is programmed is also tested separately to ensure it is valid and functional.</p><p>4. <strong>Integration and system testing: </strong>In this stage every single component and unit that is tested is then integrated, integration errors that occurred are fixed and the entire system is tested thoroughly before its deployment. Due to version control a better approach to system integration called continuous integration came into play.</p><p><strong>5. Operations and maintenance:</strong> This is the final stage of the development cycle, which involves, using the developed system practically. The process lasts for as long as the process is operational and therefore is the longest stage in the process. The maintenance of a system could involve bug fixing, code reviews, improving system architecture and improving the system as new requirements are discovered.</p><p>Every one of these five stages are very important to the classic waterfall model, but due to its lack of support to change, prototyping has been added to the entire process, in order to give customers a physical representation of the solution that is being built and to get feedback that could be vital to customer requirements. Prototyping however has unfortunately created a problem that was a result of evolutionary prototyping. Evolutionary prototyping problem occurs due to poorly made systems that are created as prototypes without, attributes, like robustness, maintainability and reliability. These systems are then built on and act as a foundation of the developed system therefore affecting the quality of the overall system.</p><p>It is disappointing however that the waterfall model is the most preferred model, with a third of developers using it in their workplaces despite its inefficiency. This inefficiency is caused by a lack of a systematic approach to modelling and analysis. In an attempt to eliminate these cases, specialists have suggested the use of different principles that can be applied with regards to the waterfall model. Some of these principles include “throw-away-prototypes”, which means to ditch prototypes that have been created, and recreate the entire system from scratch.</p><p>The waterfall model is unable to handle changing requirements. When sufficient changes occur, there is need to pause parts of the development process, and re-visit an earlier developmental stage to fix faults. When this stage is re-visited, time and budget might force the re-specified requirements to be left unfinished. The team could return to the previous stage to continue the development process in an attempt to finish on time and within budget as production and approval of the documents cost money. Requirements pausing/ freezing have an adverse effect in the sense that customer requirements could change and so the system requirement changes as well.</p><p>The waterfall model only holds a major advantage in cases where there is need for a consistent plan and none changing customer requirements. Software systems that are best created with this methodology include operating systems. According to the survey performed by abysoft.com in November to December 2013, the waterfall model had a success rate of about 49% (in cases where complete systems were delivered), challenging rate of about 33% (cases where systems were delivered but with missing functional requirements) and a failure rate of about 18%(cases where the system was not delivered).</p><h3>Lean</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/224/1*nAbf9E9QSyTNIGGK7Ytbig.png" /><figcaption>(leanAccountants, 2015)</figcaption></figure><p>Lean is not just a software development model, it involves a complete change of reasoning, analysis and work flow in the work place. Every lean culture tries its uttermost to eliminate every “just in case” process by implementing a pull-over-push method. The question then is why “pull”?</p><p>Pull implies that every functionality that is created is actually needed in the market at that point in time. Then the other question then becomes ‘why not “push”?’ implies a counter motive against forcing those functionalities that are in line with predetermined production expectations. Lean practices follows the basic work-flow of learn, build and measure.</p><p>Imagine a population of people living in the rain forests of Africa practicing lean as a culture. As you know each individuals that is part of lean will have a strong sense of believe in the shared values and principles that lean promotes. In this spirit, every development process that is lean therefore has to follow seven core principles that declares it as a member of the lean tribe. Although these principles have been rephrased and rearranged over time, below are the principles with respect to (Poppendieck and Cusumano, 2012).</p><p>1. <strong>Optimize the whole</strong>: is about discovering what a customer would like done and how this job might be done. You try your best to create the system with as much value as possible, you value criteria is completely dependent on the customer, and so you optimise development for that specific customer.</p><p>2. <strong>Eliminate waste</strong>: Eliminating waste is a very core principle of lean. As a human eating more than you need to gets you fat and eating just as much as you need to leaves you lean.</p><p>What is exactly is waste in terms of software development?</p><p>According to Mary Poppendieck from MIT, waste is anything that does not either add customer value directly or add knowledge about how to deliver that value effectively. This means that anything that isn’t beneficiary to the customer’s value is waste (otherwise known as muda). There are different type of wastes that can creep in when a system is being developed. For example loss of knowledge, unfinished work, shabby planning, and the amount of time wasted in finding defect / errors and fixing them.</p><p>3. <strong>Build Quality In</strong>: The best way to ensure quality is built in the system is by the use of continuous integration. Continuous integration occurs when individuals in a development team integrate their work frequently, in so doing eliminating the need for a gigantic integrations at the end of a development process (Fowler, 2006). The use of version control allows teams to be able to backup, recover, and view a complete sequence of baseline (a mainline). Each baseline states the collection of different components which work together to create a functional system</p><p>This means that there is immediate integration of system units that are built allowing for a working version to be deployed at the end of a baseline and then instantly tested for errors and defects. This doesn’t only reduce the amount of time spent on integration but in cases where version control is used, it provides a good platform for backup and recovery.</p><p>4. <strong>Learn constantly</strong>: is an approach that helps us to build knowledge into a system, but not just any knowledge. We build knowledge that allows us to create greater value to the customer, a better product on delivery and that helps to continuously improve service. Learn constantly involves creating a system that does the barest minimum and then performing delivery frequently in order to get feedback from a customer’s experience. This continuous learning, leads to continuous improvement (kaizen) of services. ”Learn constantly” is only one of two ways of creating knowledge that is supported by lean the other “Learn first” involves leaving decision to the last minute and at that point enough information has been gathered about the subject and a better decision can be made.</p><p>5. <strong>Deliver fast</strong>: delivering fast does not just involve delivering working software one batch after another which is actually continuous deployment. Delivering fast involves continuous delivery, this means that working systems delivered to the customer is not just working but is a complete release of the system, therefore creating a quick flow in the production process. A great support for continuous delivery is continuous integration. Continuous integration supports continuous delivery by allowing it to allowing it to be repeatable, reliable and predictable helping a team to save time, and money (Shahin, 2016).</p><p>Continuous integration therefore allows teams to be empowered by providing them a platform of self-service, helps to reduce errors in a systems configuration, lower’s stress of team members associated with the release, makes deployment flexible and makes team perfect deployment(i.e. Practice makes perfect(Shahin, 2016).</p><p>6. <strong>Energize workers</strong>: It is important for energy to remain in the work environment and so the introduction of bright young minds into the environment to bring life and optimism and dreams is highly recommended. It is important that everyone in the environment is motivated about the work place. Motivation can be brought about by having a clear purpose, facing a challenge and having a fair sense of responsibility within the group</p><p>7. <strong>Keep getting better</strong>: Continuous Improvement is very important, because there is need for self-improvement in today’s society. Insufficient improvement will leave a person, or in the case of lean a team, stuck on the same level unable to advance and therefore unable to meet a customer’s value expectations. It is therefore important to change as fast as the world changes, paying attention to the little thing. This is because they can make a big difference, and use scientific methods to test improvement.</p><p>Due to lean’s constant improvement principle, lean believes that there is no such thing as best practices, but rather good practices which are always changing as the world advances. An example of a good practice is code reviews. Unlike the waterfall model lean is a flexible, light-weight, and iterative methodology without a clear, rigid structure of what should be done. It is rather an entire framework filled with principles that can be followed in order to transform development processes. Lean’s primary goal is to eliminate waste, encourage systematic reuse of product and platforms, reduces hours spent on overlapping activities and focus on value for both the customer and the product.</p><p>These key attributes make lean a solid foundation for other software development methodologies like agile and devOps. According to the survey performed by abysoft.com in November to December 2013, lean practices had a success rate of about 72% (in cases where complete systems were delivered), challenging rate of about 20% (cases where system were delivered but with missing functional requirements) and a failure rate of about 8%(cases where the system were not delivered).</p><h3>The agile process</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/300/1*GeJwWp6Jr39GBbq92qm8zA.png" /><figcaption>(varunm.com, 2015)</figcaption></figure><p>Due to the inefficiency of the waterfall model there was need for a development methodology that could respond to changing customer requirements. In 2001, a summit was held and 17 software professionals came together to discuss and share thoughts about a solution to this problem, and agile was formed.</p><p>What is Agile?</p><p>Agile software development, otherwise known as ASD, is a lightweight software engineering methodology that is based on short cycles and iterative delivery, individual interactions and customer collaborations. ASD therefore thrives in changing customer requirements in comparison to the waterfall model. To make this happen, agile is built on certain lean principles making it lightweight. The agile manifesto that was conscripted (fowler and Highsmith, 2001; Fowler, 2002) is described below:</p><p><em>We are uncovering better ways of developing software by doing it, and helping others do it. Through this work we have come to value:</em></p><p>· <em>Individuals and interactions over processes and tools.</em></p><p>· <em>Working software over comprehensive documentation.</em></p><p>· <em>Customer collaboration over contract negotiation.</em></p><p>· <em>Responding to change over following a plan.</em></p><p><em>That is, while there is value in the items on the right, we value the items on the left more</em></p><p><em>(Agile Manifesto).</em></p><p>The agile manifesto therefore begat certain principles that must be followed for all participants of agile. There are 12 principles of agile which are extracted from (Misra et al., 2012):</p><p>1. Highest priority is given to satisfying the customer through early and continuous delivery of valuable software — This means that a customer’s satisfaction with the product that will be made available to them before hand in made the highest priority.</p><p>2. Changing requirements are welcome, even late in development. Agile processes harness change for the customer’s competitive advantage — This means that agile encourages a customer’s changing requirements allowing customer to keep up with present trends with regards to products therefore granting them a competitive advantage, unlike the waterfall model that creates the initial specification and not making room for changing customer requirements.</p><p>3. Working software is delivered frequently, from a couple of weeks to a couple of months, with a preference to the shorter timescale — Agile promotes working software and so continuous delivery which means that continuous integration is important, here it is easy to see how agile models lean behaviours. The shorter time-scale means customers can gauge progress of the team.</p><p>4. Business people and developers work together daily throughout the project — This means that the created system will be completely inline with the customer requirements.</p><p>5. Projects are built around motivated individuals — This principle is built on the lean principle “Energize workers” and highlights the importance of having happy people on development teams. Human beings build systems and human being are motivated individuals therefore motivation plays a vital role in the creation of software.</p><p>6. It is believed that the most efficient and effective method of conveying information to and within a development team is face-to-face conversation: Agile believes that face to face communication is the best means of communication in a development.</p><p>7. It is believed that working software is the primary measure of progress: Agile believes that a software that does all its functional requirement adequately, i.e. verified and valid is how software is measured best.</p><p>8. It is believed that agile processes promote sustainable development. The sponsors, developers, and users should be able to maintain a constant pace indefinitely.</p><p>9. It is believed that continuous attention to technical excellence and good design enhances agility: This principle is derived from the lean principle “keep getting better” in the sense that there is a need to pay attention to detail in the system in order to ensure technical excellence, Technical excellence is also dependant on the technological advancement made (i.e. keeping up with the season).</p><p>10. It is believed that simplicity is essential — This principle is derived from lean’s eliminating waste because it stands against unnecessary functionality that doesn’t add value to the system meaning simplicity is lean.</p><p>11. It is believed that the best architectures, requirements, and designs emerge from self-organizing teams: Agile ensures that teams are empowered by giving key roles to individuals in teams who are closer to one another, have face-to-face conversations and organisation and distribution of work is made easier as everyone understands the scope of the system.</p><p>12. It is believed that success is achieved when at regular intervals the team reflects on how to become more effective, then tunes and adjusts its behaviour accordingly: Agile believes strongly in continuous improvement.</p><p>These agile principles are very core to the approach of ASD, the main aspects of the concepts are its lightweight nature, small plans of a feature to be developed, its iterative development, continuous delivery and the collaborations it makes possible. There are many different application of ASD which has a bit of uniqueness but all under the agile culture, some of which are Crystal Family of Methodologies (CM), eXtreme Programming Development Process (XP), Scrum Development Process, and Dynamic System Development Method (DSDM). Despite their uniqueness the standard ASD methods and processes range between three to five stages (Goeschka and Radinger, 2003):</p><p>- <strong>Requirements engineering</strong>: which involves the customer and contains a detailed plan and system features.</p><p>- <strong>Planning</strong>: A plan and implementation schedule based on the desired feature.</p><p>- <strong>Design</strong>: The developers build the design model.</p><p>- Build: The implementation occurs and based on the design.</p><p>- <strong>Test and release</strong> — The build is integrated, tested, documented and released to be viewed by the customer.</p><p>Another important thing to note is the team size of the group, in order for individuals and interactions to be valued, a group has a maximum of ten people. According to the survey performed by abysoft.com in November to December 2013, ASD had a success rate of about 62% (in cases where complete system were delivered), challenging rate of about 34% (cases where systems was delivered but with missing functional requirements) and a failure rate of about 4%(cases where the systems was not delivered).</p><h3>DevOps</h3><figure><img alt="" src="https://cdn-images-1.medium.com/max/601/1*d_iv1Rxoye2u4go2wMVpfA.png" /><figcaption>(Wurster, 2013)</figcaption></figure><p>DevOps is a combined word for development and operation. It is a new software development framework that is used to build large-scale organisational system. It is a cultural and professional movement, focused on how we build and operate high velocity organisation born from the experiences of its practitioners (A. Jacobs, 2013).</p><p>DevOps is not just a development process but a cultural shift. DevOps brings a culture to focus in an organisation attempting to solve cultural issues, make system development processes more liquid by breaking down the barriers between IT developers and operations expert. This is done by helping them working hand in hand in order to produce good software.</p><p>The devOps culture was brought about as a result of wasted time spent on integration, testing and deployment of code written by the developers and then lengthy periods that IT operations teams take to move written code into its production phase. DevOps therefore provides a paradigm shift that allows the developers and operation experts to understand enough about the subject matter and each other in so doing improving collaboration and relationships between teams. This quickens the implementation and deployment process (with the use of automation) making software delivery quickly available to the customers.</p><p>DevOps mix of a bunch of methodology and so it does not have set principles. Although there are certain practices that are agreed upon, as with every culture, some agreed upon practices are as follows (Wurster, 2013):</p><p>- <strong>Cross-functional teams and skills</strong>: means that the team is made up of individuals with core skills that are of importance to the present project.</p><p>- <strong>Continuous delivery</strong>: means that software is released in small batches. As with agile and lean.</p><p>- <strong>Continuous assessment</strong>: means that teams will continually assess performance in order to discover faults and continuously improve as with agile and lean.</p><p>- <strong>Optimization of toolsets</strong>: the use of the same type of tool for each member in the team regardless of their speciality when it is possible.</p><p>- <strong>Automated deployment pipeline</strong>: means the process of moving code from implementation to production will be as automated as possible therefore saving time and therefore automating work-flow and allowing for easy scale of deployment.</p><p>DevOps lean culture allows its practitioners to continuously improve by assessing their performances iteratively. It acts on agile principles by requiring iterative releases of a system, in so doing, forces a need for continuous deployment and integration. Continuous integration is however a key aspect with all kinds of software development processes people of its ability to help backup, recover, store versions, wide-access, source control, version management, configuration management and many more.</p><p>The wide variety of fields involved in devOps include lean, agile, continuous improvement, social psychology, ITIL (IT Infrastructure Library), Resilient (includes Recovery orientated computing) and Systems thinking. Each of these fields involves crucial aspects about devOps.</p><h3>Conclusion</h3><p>In conclusion, it is important to realise that a software model or methodology that does not cater for change will fail often. This is because change is a part of the development process and not an error or a bug that needs to be fixed. A development team that copes with change through collaboration and communication with a customer regularly is therefore gathering a better knowledge of the deployed environment. When sufficient improvement metrics are met periodically, success is guaranteed. DevOps is not just lean but includes agile, it is a complete shift in culture that when practiced correctly it becomes a contagious culture that can change the way things are done across an organisation, optimising development processes and people’s attitudes.</p><h3>References</h3><p>1. Ambysoft.com. (2016). <em>2013 IT Project Success Rates Survey Results</em>. [online] Available at: <a href="http://www.ambysoft.com/surveys/success2013.html">http://www.ambysoft.com/surveys/success2013.html</a> .</p><p>2. Birchall, H. and B a u e r, R. (2016). M?iNAGXNG LARGE SCALE SOFTWARE DEVELOPMENT WITH AN AUTOMATED CHANGE CONTROL SYSTEM.<em>1</em>, 1(1), p.All.</p><p>3. Davis, A. and Sitaram, P. (1994). A concurrent process model of software development. <em>SIGSOFT Softw. Eng. Notes</em>, 19(2), pp.38–51.</p><p>4. Fowler, M. (2006). <em>Continuous Integration</em>. [online] martinfowler.com. Available at: <a href="http://www.martinfowler.com/articles/continuousIntegration.html">http://www.martinfowler.com/articles/continuousIntegration.html</a></p><p>5. Fowler, M. (2016). Continuous Delivery. <em>1</em>, 1(1), pp.17–22.</p><p>6. Goeschka, K. and Radinger, W. (2003). Agile Software Development for Component Based Software Engineering. <em>1</em>, 1(1), p.all.</p><p>7. Habib, M. (2013). <em>Agile software development methodologies and how to apply them — CodeProject</em>. [online] Codeproject.com. Available at: <a href="http://www.codeproject.com/Articles/604417/Agile-software-development-methodologies-and-how-t">http://www.codeproject.com/Articles/604417/Agile-software-development-methodologies-and-how-t</a> .</p><p>8. Itinfo.am. (2016). <em>Software Development Methodologies</em>. [online] Available at: <a href="http://www.itinfo.am/eng/software-development-methodologies/">http://www.itinfo.am/eng/software-development-methodologies/</a> .</p><p>9. Karibskii, A. (1991). Managing the development of large-scale systems. <em>Mathematics and Computers in Simulation</em>, 33(4), pp.287–293.</p><p>10. Mahdi, A., Hamed, M. and Abushama, H. (2013). Popular Agile Approaches in Software Development: Review and Analysis. <em>1</em>, 1(1), p.All.</p><p>11. Msdn.microsoft.com. (2016). <em>Agile Principles and Values, by Jeff Sutherland</em>. [online] Available at: <a href="https://msdn.microsoft.com/en-us/library/dd997578.aspx.">https://msdn.microsoft.com/en-us/library/dd997578.aspx.</a></p><p>12. Phillip, A., Laplante, Colin, J. and Neill, J. (2016). “The Demise of the Waterfall Model Is Imminent” and Other Urban Myths. <em>1</em>, 1(1), p.All.</p><p>13. Poppendieck, M. (2007). Lean Software Development. <em>1</em>, [online] 1(1), p.All. Available at: <a href="http://ieee.org./">http://ieee.org.</a></p><p>14. Poppendieck, M. and Cusumano, M. (2012). Lean Software Development: A Tutorial.<em>IEEE Softw.</em>, 29(5), pp.26–32.</p><p>15. Processimpact.com. (2016). <em>Creating a Software Engineering Culture</em>. [online] Available at: <a href="http://www.processimpact.com/articles/culture.html">http://www.processimpact.com/articles/culture.html</a></p><p>16. Roche, J. (2013). Adopting DevOps practices in quality assurance. <em>Communications of the ACM</em>, 56(11), pp.38–43.</p><p>17. Shahin, M. (2016). Architecting for DevOps and Continuous Deployment. <em>1</em>, 1(1), p.All.</p><p>18. Wang, X., Conboy, K. and Cawley, O. (2012). “Leagile” software development: An experience report analysis of the application of lean approaches in agile software development. <em>Journal of Systems and Software</em>, 85(6), pp.1287–1299.</p><p>19. Zhu, L., Bass, L. and Champlin-Scharff, G. (2016). DevOps and Its Practices. <em>IEEE Softw.</em>, 33(3), pp.32–34.</p><p>20. Wurster, (2013). Emerging Technology Analysis: DevOps a Culture Shift, Not a Technology. <em>1</em>, 1(1), p.All.</p><p>21. Sommerville, I. and Stevens, P. (2012). <em>Introduction to software engineering</em>. Frenchs Forest, N.S.W.: Pearson.</p><p>22. Misra, S., Kumar, V., Kumar, U., Fantazy, K. and Akhter, M. (2012). Agile software development practices: evolution, principles, and criticisms. <em>Int J Qual &amp; Reliability Mgmt</em>, 29(9), pp.972–980.</p><img src="https://medium.com/_/stat?event=post.clientViewed&referrerSource=full_rss&postId=981ae483968a" width="1" height="1" alt="">]]></content:encoded>
        </item>
    </channel>
</rss>