Saudi Cultural Missions Theses & Dissertations
Permanent URI for this communityhttps://drepo.sdl.edu.sa/handle/20.500.14154/10
Browse
2 results
Search Results
Item Restricted Improving Feature Location in Source Code via Large Language Model-Based Descriptive Annotations(Arizona State University, 2025-05) Alneif, Sultan; Alhindawi, NouhFeature location is a crucial task in software maintenance, aiding developers in identifying the precise segments of code responsible for specific functionalities. Traditional feature location methods, such as grep and static analysis, often result in high false-positive rates and inadequate ranking accuracy, increasing developer effort and reducing productivity. Information Retrieval (IR) techniques like Latent Semantic Indexing (LSI) have improved precision and recall but still struggle with lexical mismatches and semantic ambiguities. This research introduces an innovative method to enhance feature location by augmenting source code corpora with descriptive annotations generated by Large Language Models (LLMs), specifically Code Llama. The enriched corpora provide deeper semantic contexts, improving the alignment between developer queries and relevant source code components. Empirical evaluations were conducted on two open-source systems, HippoDraw and Qt, using standard IR performance metrics: precision, recall, First Relevant Position (FRP), and Last Relevant Position (LRP). Results showed significant performance gains; a 40% precision improvement in HippoDraw, and a 26% improvement in Qt. Recall improved by 32% in HippoDraw and 24% in Qt. The findings highlight the efficacy of incorporating LLM-generated annotations, significantly reducing developer effort and enhancing software comprehension and maintainability. This research provides a practical and scalable solution for software maintenance and evolution tasks.13 0Item Restricted Dynamic Feature Location Framework for Software Project(University of Bahrain, 2024-08) Buzaid, Faisal; Albalooshi, FawziThe Dynamic Feature Location Techniques (DFLTs) aim to automate the process of identifying the source code responsible for executing specific features within software systems. Manual implementation of DFLTs is time-consuming and demanding for developers, leading to the proposal of semi-automated approaches. One common approach involves generating execution traces by executing multiple scenarios for each software feature and then mapping the corresponding source code based on these traces. However, the execution traces are often large and contain irrelevant data to the software feature, requiring solutions to reduce their size and the eliminate irrelevant data. One such solution involves minimizing the number of scenarios needed to exercise a software feature, but little work has been done in this area. To address this gap, a generic framework called Aggregation of Execution Traces to Formulate a Scenario (AETFS) is introduced in this work. AETFS leverages runtime software output and employs textual analysis techniques to extract relevant data from the execution trace for scenario creation. It explores textual analysis, including topic modeling, as a means to select accurate scenarios for DFLTs. The performance of AETFS is characterized in terms of execution trace granularity, enabling the identification of meaningful terms that can filter the execution trace using textual analysis techniques such as Latent Semantic Indexing (LSI). The evaluation encompasses eight subject systems with 600 features, making it more extensive than previous studies. The study identifies certain attributes of execution traces and text queries that impact AETFS’s performance. Two distinct groups emerge, one achieving superior Feature Location (FL) using AETFS and the other achieving better FL using a conventional baseline method. Combining AETFS with the baseline method significantly enhances performance, with the top results surpassing the baseline by 45% and the lowest by 12% over AETFS. In conclusion, this work highlights the importance of rigorously characterizing the proposed DFLTs framework to identify optimal scenarios for exercising software features. It emphasizes the need to differentiate between scenarios and their characterizations to generate necessary insights. The findings demonstrate the effectiveness of AETFS while providing valuable insights for further advancements in the field of DFLTs.14 0