A Topic Modeling Based Solution for Confirming Software Documentation Quality

Abstract

Abstract-this paper presents an approach for evaluating and confirming the quality of the external software documentation using topic modeling. Typically, the quality of the external documentation has to mirror precisely the organization of the source code. Therefore, the elements of such documentation should be strongly written, associated, and presented. In this paper, we use Latent Dirichlet Allocation (LDA) and HELLINGER DISTANCE to compute the similarities between the fragments of source code and the external documentation topics. These similarities are used in this paper to improve and advance the existing external documentation. Furthermore, these similarities can also be used for evaluating the new documenting process during the evolution phase of the software. The results show that the new approach yields state-of-the-art performance in evaluating and confirming the existing external documentations quality and superiority.

New Approach for Breast Cancer Detection- Based Machine Learning Technique

“…Figure 6 provides a comprehensive summary of breast cancer statistics across various demographics in different countries (DeSantis et al, 2019;Wilkinson & Gathani, 2022). Consequently, effective methods and tools are required for early breast cancer identification, since this is a prerequisite for effective breast cancer treatment (Alhindawi, et al, 2016;Barrios, 2022;Wilkinson & Gathani, 2022). Machine learning (ML) methods have been applied in various healthcare areas to identify BC (i.e., "Breast Cancer") in recent years.…”

Media narratives on (Ecological) sustainability: A topic modeling analysis of entrepreneurship news in the MENA region

“…For each model the value of topic coherence was established [83,84] and graphed across the number of topics (see Figs 2 and 3). The Alpha & Beta parameters were left at their standard settings of 1/#topics and 0.1 respectively [63,85]. A comparison of the coherence measure across a wide range of Alpha & Beta values confirmed these to yield the best results while there was also no visible difference when inspecting the topics and keywords.…”