Coding conventions are ubiquitous in software engineering practice. Maintaining a uniform coding style allows software development teams to communicate through code by making the code clear and, thus, readable and maintainable—two important properties of good code since developers spend the majority of their time maintaining software systems. This dissertation introduces a set of probabilistic machine learning models of source code that learn coding conventions directly from source code written in a mostly conventional style. This alleviates the coding convention enforcement problem, where conventions need to first be formulated clearly into unambiguous rules and then be coded in order to be enforced; a tedious and costly process. First, we introduce the problem of inferring a variable’s name given its usage context and address this problem by creating Naturalize — a machine learning framework that learns to suggest conventional variable names. Two machine learning models, a simple n-gram language model and a specialized neural log-bilinear context model are trained to understand the role and function of each variable and suggest new stylistically consistent variable names. The neural log-bilinear model can even suggest previously unseen names by composing them from subtokens (i.e. sub-components of code identifiers). The suggestions of the models achieve 90% accuracy when suggesting variable names at the top 20% most confident locations, rendering the suggestion system usable in practice. We then turn our attention to the significantly harder method naming problem. Learning to name methods, by looking only at the code tokens within their body, requires a good understating of the semantics of the code contained in a single method. To achieve this, we introduce a novel neural convolutional attention network that learns to generate the name of a method by sequentially predicting its subtokens. This is achieved by focusing on different parts of the code and potentially directly using body (sub)tokens even when they have never been seen before. This model achieves an F1 score of 51% on the top five suggestions when naming methods of real-world open-source projects. Learning about naming code conventions uses the syntactic structure of the code to infer names that implicitly relate to code semantics. However, syntactic similarities and differences obscure code semantics. Therefore, to capture features of semantic operations with machine learning, we need methods that learn semantic continuous logical representations. To achieve this ambitious goal, we focus our investigation on logic and algebraic symbolic expressions and design a neural equivalence network architecture that learns semantic vector representations of expressions in a syntax-driven way, while solely retaining semantics. We show that equivalence networks learn significantly better semantic vector representations compared to other, existing, neural network architectures. Finally, we present an unsupervised machine learning model for mining syntactic and semantic code idioms. Code idioms are conventional “mental chunks” of code that serve a single semantic purpose and are commonly used by practitioners. To achieve this, we employ Bayesian nonparametric inference on tree substitution grammars. We present a wide range of evidence that the resulting syntactic idioms are meaningful, demonstrating that they do indeed recur across software projects and that they occur more frequently in illustrative code examples collected from a Q&A site. These syntactic idioms can be used as a form of automatic documentation of coding practices of a programming language or an API. We also mine semantic loop idioms, i.e. highly abstracted but semantic-preserving idioms of loop operations. We show that semantic idioms provide data-driven guidance during the creation of software engineering tools by mining common semantic patterns, such as candidate refactoring locations. This gives data-based evidence to tool, API and language designers about general, domain and project-specific coding patterns, who instead of relying solely on their intuition, can use semantic idioms to achieve greater coverage of their tool or new API or language feature. We demonstrate this by creating a tool that suggests loop refactorings into functional constructs in LINQ. Semantic loop idioms also provide data-driven evidence for introducing new APIs or programming language features.

Made available in DSpace on 2014-06-11T19:34:58Z (GMT). No. of bitstreams: 0 Previous issue date: 2007-12-20Bitstream added on 2014-06-13T19:44:43Z : No. of bitstreams: 1 pastorelli_als_dr_guara.pdf: 2377495 bytes, checksum: 6616c1dd459e5d7fa4bd2985247b67f6 (MD5) / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES) / O trabalho tem por objetivo recuperar a usabilidade dos processos de ensaios em túnel de vento por meio de recursos computacionais, visando a atualização e revitalização do sistema de software através de recursos tecnológicos, interfaces, aplicativos de apoio avançados e da implantação de práticas de engenharia e qualidade de software. Neste trabalho de tese fez-se estudo de práticas, métodos, processos e técnicas de engenharia de software de modo a aplicá-las adequadamente no ciclo de desenvolvimento do projeto do sistema de software, considerando o conceito atual de software que abrange os programas de computadores, a documentação associada e os dados necessários para que os softwares de aquisição e redução de dados operem corretamente e continuamente durante a modernização do sistema de medidas do túnel de vento. Foi dado enfoque na melhoria da qualidade dos produtos e dos processos gerenciais e técnicos, considerando as ações das metodologias tradicionais e ágeis adequadas ao escopo e ao grupo de projeto. Neste trabalho, para atingir a abrangência do conceito de software e manter a funcionalidade do sistema de software durante a modernização do túnel de vento é gerada a metodologia para aplicação de reuso e reengenharia nos software de aquisição e redução de dados de um túnel de vento, denominada Metodologia Especial Tradicional e Ágil de aplicação em Projeto Legado de software em Inovação – METAPLI. Com a implantação desta metodologia de projeto de engenharia de software se provê a organização dos processos gerenciais e técnicos, se gera a documentação de todo projeto do sistema de software e se preserva o conhecimento dos processos de engenharia. / This work focuses on recovering the usability of the wind tunnel tests processes by means of computational resources in order to updating and revitalizing the software system through the technological resources, interfaces, advanced applicative of the aid and the establishing of software engineering and software quality practices. On this thesis’ work they were performed studies of practices, methods, processes and software engineering techniques in a way of applying them adequately on the project development cycle of the software system, considering the software current concept which recovers the computer programs, the associated documentation and the necessary data in a way that the data acquisition and reduction software operates correctly and continuously along the wind tunnel measurement system’s modernization. The focus was on quality improvement of products and managerial and technical processes, considering the traditional and agile methodologies actions, which are adequate to scope and to project group. On this work, to achieve the software concept wideness, keeping the software system functionality along the wind tunnel modernization, it was generated the methodology in order to apply the reuse and reengineering of data acquisition and reduction software of a wind tunnel, which is denominated “Traditional and Agile Special Methodology for application in Project Legacy of software in Innovation - METAPLI”. The implantation of this project methodology of software engineering proves the managerial and technical processes organization, generates the documentation of the whole software system project and preserves the knowledge of engineering processes.

With a software industry that is moving at a fast pace, continuous integration and delivery is something important for many products today. Moreover, with containers being on the rise since 2013, more companies are moving their CI/CD environment into containers not only for development but also for testing. This thesis begins with giving the reader an introduction to containers, container orchestration, and Jenkins, which is a continuous integration and delivery tool. The experiment was then set up with one container based cluster and one single node machine. Two kinds of experiments were run on them, one big job and one small job. The system scalability is assessed, and with smaller clusters the memory overhead could be an issue. Performance wise, the container cluster is performing better than a single node machine, as long as it is utilizing all its nodes. Security with containers is still an issue and it could be fatal for a cluster if it is compromised.

Det denna rapport kommer att avhandla är ett projekt genomfört hos Roxtec AB, syftet har i detta projekt varit att underlätta hanteringen av så kallade utm-adresser. Vi kommer att gå igenom problemställning, hur vi väljer att lösa samma problem. Senare kommer vi att behandla hur vi löste de här problemen, huvudsyftet har varit att skapa ett enkelt verktyg för att generera nyss nämnda adresser. Lite extrafunktioner runt de här funktionerna kommer att avhandlas såsom sparning av historik, inlogg och så vidare. Projektet kom att innehålla många rader kod, de här kommer inte rapporten att avhandla i detalj, förutom vissa mer betydelsefulla delar av projektet. I de mer betydelsefulla avsnittet går vi igenom dess funktion samt med illustrationer försöker vi förklara dess syfte på ett enkelt sätt. Hur väl vi lyckats att lösa de uppställda problemen vi ställs inför går vi igenom och diskuterar även de här till sist. Projektet kommer att skapas i .NET och i lämplig programvara för att utföra och lösa senare uppställd problematik.

No description available.

I dagens webb används ofta många olika innehållshanteringssystem för att göra det enkelt för användare att hantera deras egna hemsidor. I utvecklingsarbetet hos webbyråer bedrivs även någon typ av projektledning. Syftet med denna undersökning är att utvärdera och analysera ett företags befintliga CMS och deras arbetsprocess. Denna utvärderingen kommer göras utifrån ett visst antal kriterier som är kännetecken för olika projektledningsmetoder samt en separat utvärdering för ett passande CMS utifrån specifikationer och intervjuer. Här kommer faktorer som licens, funktionalitet, plattformsstöd, vidareutveckling, popularitet samt exempelvis programmeringsspråk. Denna undersökning visade på att företaget arbetar med ett agilt arbetssätt, samt att deras befintliga CMsystem är det optimala för deras verksamhet för att de ska upprätthålla sina möjligheter att helt skräddarsy lösningar och ha full kontroll på produkten.

Projektet avser att implementera spelifieringselement på den ideella bloggen 400contacts, en blogg som coachar och tipsar om jobbsökning i Sverige för nyanlända. Detta ska uppnås genom att använda sig av modeller av Yu-kai Chou, en framstående person inom området om spelifiering. Målet var till en början att öka antalet återkommande besökare till 50% av alla besökare men ändrades senare i projektet till att konvertera besökare till att ansluta sig till företagets facebookgrupp. För att lyckas uppnå detta har en ny design till bloggen tagits fram och under arbetets gång har ett poängsystem tagits fram som styr belöningar i form av märken på användarnas profiler. Dessa poäng ställer även användarna i en topplista. Poäng och märken förtjänas genom att vara delaktig på sidan, exempelvis genom att logga in eller att läsa ett nytt blogginlägg. Webbplatsen är byggd i WordPress och förlitar sig kraftigt på de tillgängliga tillägg som plattformen erbjuder. Arbetet har inte stött på större problem och har lyckats uppnå båda mål i sitt slutresultat.

AbstractMachine learning is a subfield within artificial intelligence that revolves around constructingalgorithms that can learn from, and make predictions on data. Instead of following strict andstatic instruction, the system operates by adapting and learning from input data in order tomake predictions and decisions. This work will focus on a subcategory of machine learningcalled “MultilabelClassification”, which is the concept of where items introduced to thesystem is categorized by an analytical model, learned through supervised learning, whereeach instance of the dataset can belong to multiple labels, or classes.This paper presents the task of implementing a multilabelclassifier based on the ID3algorithm, which we call MLID (MultilabelIterative Dichotomiser). The solution is presentedboth in a sequentially executed version as well as an parallelized one.We also presents acomparison based on accuracy and execution time, that is performed against algorithms of asimilar nature in order to evaluate the viability of using ID3 as a base to further expand andbuild upon in regards of multi label classification.In order to evaluate the performance of the MLID algorithm, we have measured theexecution time, accuracy, and made a summarization of precision and recall into what iscalled Fmeasure,which is the harmonic mean of both precision and sensitivity of thealgorithm. These results are then compared to already defined and established algorithms,on a range of datasets of varying sizes, in order to assess the viability of the MLID algorithm.The results produced when comparing MLID against other multilabelalgorithms such asBinary relevance, Classifier Chains and Random Trees shows that MLID can compete withother classifiers in term of accuracy and Fmeasure,but in terms of training the algorithm,the time required is proven inferior. Through these results, we can conclude that MLID is aviable option to use as a multilabelclassifier. Although, some constraints inherited from theoriginal ID3 algorithm does impede the full utility of the algorithm, we are certain thatfollowing the same path of development and improvement as ID3 experienced would allowMLID to develop towards a suitable choice of algorithm for a diverse range of multilabelclassification problems.

With the recent growth of the mobile market, companies want to target mobile devices while at the same time keeping product development costs low. One way to do this is to develop web applications, which are accessed from a mobile de- vice’s web browser, instead of native applications. The same web application can then be used on different platforms such as Android and iOS. However, devices such as smart phones and tablets often include cameras and sensors that a web ap- plication may want to access, but which are only accessible from native applica- tions. A framework was developed that enables web applications to communicate with native Android applications. Native applications are launched by clicking a link in the browser, and the result produced is made available to the web applica- tion through a HTTP POST request or a local web server running on the device. Key characteristics of the framework include ease of extension and the ability to enable secure (SSL) communication if desired. The ZXing Barcode Scanner ap- plication was integrated with the framework so that a scanned barcode can be dis- played in the browser. Performance measurements were conducted measuring the time taken from clicking a link to start a test application to the result being avail- able in the browser. The mean times measured were between 323 and 394 mil- liseconds. This indicates that the method used is sufficiently fast to not detract from the user experience. Future work could expand on the measurements or per- form a feature and performance comparison with PhoneGap.

M.B.A. / The aim of this research was to determine why information technology software projects fail in South Africa in order to determine whether there is a cause and cost similarity between international first world and South African IT Software Project failures. This research topic was primarily chosen because there is a lack of research on IT Software project failure in South Africa. The following objectives were defined to support the aim of the research: To ascertain project cost failure statistics sources internationally from relevant literature. To ascertain from literature what has been said about causes/reasons of project failure internationally. To ascertain South African IT Software Project failure statistics and reasons by means of a questionnaire. To determine whether the literature on international IT software project failure and South African IT software project failure compare. The research project was conducted amongst members of the Project Management Institute of South Africa (PMISA). In addition, companies that are not members of PMISA were also contacted and requested to respond to an e-mail questionnaire. It is interesting to note that one of the reasons for projects being challenged and cancelled, deals with, or is related to, requirements. Changing and / unclear user requirements was always listed in the top three reasons for projects being challenged and cancelled. The top three reasons listed by The Standish Group also lists changing and / unclear user requirements in the top three. This research would therefore suggest that effectively dealing with changing and / unclear user requirements would significantly increase project success. In conclusion, the comparison between South African companies and International companies revealed that the causes are for the most part the same, and the costs appear to be vastly different, but similar in trend. As is evident from the research literature, this research topic rarely studied, and therefore additional research can be done to explore this topic.