Abstract To address the lexical gaps between natural language (NL) queries and Application Programming Interface (API) documentations, and between NL queries and programme code, this study developed a novel approach for recommending Java API classes that are relevant to the program​ming tasks described in NL queries. A Doc2Vec model was trained using question titles mined from Stack Overflow. The model was used to find question titles that are semantically similar to a query. Latent Dirichlet Allocation (LDA) topic modelling was applied on the Java API classes (extracted from code snippets found in the accepted answers of these similar questions) to extract a single topic comprising of the Top‐10 Java API classes that are relevant to the query. The benchmarking of the proposed approach against state‐of‐the‐art approaches, RACK and NLP2API, by using four performance metrics show that it is possible to produce comparable API recommendation results using a less complex approach that makes use of some basic machine learning models, in particular, Doc2Vec and LDA. The approach was implemented in a Java API class recommender with an Eclipse IDE's plug‐in serving as the front‐end.

Dynamic software updating (DSU) is shifting gears to modify software systems without a halt. Even though extensive research has been conducted on DSU, it is necessary to synthesise and map the results of recent studies on DSU for prospective research highlights. This study aims to highlight the current state‐of‐the‐art, to recognise trends, and to identify existing open issues in DSU. A systematic mapping study (SMS) was conducted with a set of six research questions. A total of 1066 papers published from 2005 to 2019 were recorded. After a filtering process, 112 primary studies were selected and inspected. This study highlights the current state‐of‐the‐art of DSU including approaches, tools, models, and techniques. Also, this study outlines application domains, research type, and contributions type facets in DSU. In addition, this study demonstrates benchmarks, datasets, evaluation metrics, and existing open issues for future research in DSU. The results of this investigation can be used to highlight current state‐of‐the‐art of DSU, to show trends of DSU, and to demonstrate existing open issues in DSU.

Navigating and understanding complex software architecture documentation is often challenging. To support finding relevant information in architecture documents (ADs), we propose a semi-automated approach based on the actual usage of ADs by previous users, i.e. by capturing users' exploration paths through ADs and making these paths available for future retracing and analysis. To do this, we have built a prototype tool (KaitoroCap) that captures users' AD exploration paths and saves them with contextual metadata. KaitoroCap displays the exploration paths in hierarchical tree views and these exploration paths can be searched. This is helpful for recalling previous navigations and to follow others' useful paths in finding relevant information in AD. Our approach also enables dynamic restructuring of ADs and incorporates user rating, tagging and commenting of the content of ADs. Initial user evaluation shows promising results.

To facilitate the usage of software architecture documents (ADs), we claim the architectural information in the ADs needs to be structured into or presented as chunks. A chunk allows related information to be retrieved collectively as a unit and simplifies information location tasks. We propose a new semi-automated approach based on the actual usage of ADs by previous users, i.e. by capturing users' exploration paths through ADs while engaging in information seeking tasks and making these paths available for future retracing and analysis. As part of our work, we developed KaitoroCap, a document navigation capture and visualisation tool. Its main features are exploration paths capture, retrieval, analysis, hierarchical tree-view visualization of paths, path searching, section rating, tagging, commenting, expanding/collapsing and page model generation to enable dynamic restructuring of ADs. This paper describes the design, implementation and usage examples of KaitoroCap.

This paper describes a software architecture documentation tool (KaitoroBase) built within the Thinkbase Visual Wiki to provide support for non-linear navigation and visualization of Software Architecture Documents (SADs) produced using the Attribute-Driven Design (ADD) method. This involves constructing the meta-model for the SAD in Freebase which provides the foundation for the graph-based interactive visualization enabled by Thinkbase. The resulting tool displays a graphical, high-level structure of SAD, allows for exploratory search, non-linear navigation, and at the same time connects to low-level details of SADs in a wiki.