Software repository hosting services contain large amounts of open-source software, with GitHub hosting over 200 million repositories, from new to established ones. However, these repositories are not easy to find, calling for various attempts to classify their application domains automatically. However, most proposed approaches use artifacts, like README files, as a proxy for the project, losing the information in the source code and the interaction between files. Furthermore, they all focus on the project-level, ignoring the decomposition of software projects into components and modules.This work presents a weak labelling approach based on keyword extraction to annotate source files in a software project.Our findings suggest that using keywords to perform file-level annotations is an effective approach that can capture enough information from the source file so that new labels can be predicted.The long-term goal of our research is to classify source code files and use these annotations to identify semantic components in software projects. In addition, these annotations can be used for semantic reverse engineering, software reuse, and more. We plan to train machine learning models that use our proposed weak supervision to better annotate source files inside software projects.

Abstract Context GitHub is the world's most prominent host of source code, with more than 327M repositories. However, most of these repositories are not labelled or inadequately, making it harder for users to find relevant projects. Various proposals for software application domain classification over the past years have been proposed. However, these several of those approaches suffer from multiple issues, called antipatterns of software classification, that reduce their usability. Objective In this paper, we propose a new taxonomy in the GitHub ecosystem, called GitRanking, starting from a well‐structured data set, composed of curated repositories annotated with topics. The main objective is to create a baseline methodology for software classification that is expandable, hierarchical, grounded in a knowledge base, and free of antipatterns. Method We collected 121K topics from GitHub and used GitRanking to create a taxonomy of 301 ranked application domains. GitRanking (1) uses active sampling to ensure a minimal number of annotations to create the ranking; and (2) links each topic to Wikidata, reducing ambiguities and improving the reusability of the taxonomy. Furthermore, we adopt the conceived taxonomy in a classification task by considering a state‐of‐the‐art classifier. Results Our results show that GitRanking can effectively rank terms in a hierarchy according to how general or specific their meaning is. Furthermore, we show that GitRanking is a dynamically extensible method: it can currently accept further terms to be ranked, and with a minimum number of annotations (). Concerning the classification task, we show that the model achieves an F1‐score of 34%, with a precision of 54%. Conclusion This paper is the first collective attempt at building a ground‐up taxonomy of software domains. Our vision is that our taxonomy, and its extensibility, can be used to better and more precisely label software projects.

Empirical results in software engineering have long started to show that findings are unlikely to be applicable to all software systems, or any domain: results need to be evaluated in specified contexts, and limited to the type of systems that they were extracted from. This is a known issue, and requires the establishment of a classification of software types. This paper makes two contributions: the first is to evaluate the quality of the current software classifications landscape. The second is to perform a case study showing how to create a classification of software types using a curated set of software systems. Our contributions show that existing, and very likely even new, classification attempts are deemed to fail for one or more issues, that we named as the ‘antipatterns’ of software classification tasks. We collected 7 of these antipatterns that emerge from both our case study, and the existing classifications. These antipatterns represent recurring issues in a classification, so we discuss practical ways to help researchers avoid these pitfalls. It becomes clear that classification attempts must also face the daunting task of formulating a taxonomy of software types, with the objective of establishing a hierarchy of categories in a classification.

Component Based Software Engineering (CBSE) seeks to promote the reuse of software by using existing software modules into the development process. However, the availability of such a reusable component is not immediate and is costly and time consuming. As an alternative, the extraction from preexisting OO software can be considered.In this work, we evaluate two community detection algorithms for the task of software components identification. Considering `components' as `communities', the aim is to evaluate how independent, yet cohesive, the components are when extracted by structurally informed algorithms.We analyze 412 Java systems and evaluate the cohesion of the extracted communities using four document representation techniques. The evaluation aims to find which algorithm extracts the most semantically cohesive, yet separated communities.The results show a good performance in both algorithms, however, each has its own strengths. Leiden extracts less cohesive, but better separated, and better clustered components that depend more on similar ones. Infomap, on the other side, creates more cohesive, slightly overlapping clusters that are less likely to depend on other semantically similar components.

Frame-semantic annotations exist for a tiny fraction of the world’s languages, Wikidata, however, links knowledge base triples to texts in many languages, providing a common, distant supervision signal for semantic parsers. We present WikiBank, a multilingual resource of partial semantic structures that can be used to extend pre-existing resources rather than creating new man-made resources from scratch. We also integrate this form of supervision into an off-the-shelf frame-semantic parser and allow cross-lingual transfer. Using Google’s Sling architecture, we show significant improvements on the English and Spanish CoNLL 2009 datasets, whether training on the full available datasets or small subsamples thereof.