Information and Software Technology最新文献

{"title":"Towards an understanding of requirements management in software ecosystems","authors":"Paulo Malcher ,&nbsp;Davi Viana ,&nbsp;Pablo Oliveira Antonino ,&nbsp;Rodrigo Pereira dos Santos","doi":"10.1016/j.infsof.2025.107672","DOIUrl":"10.1016/j.infsof.2025.107672","url":null,"abstract":"<div><h3>Context:</h3><div>Software ecosystems (SECO) have introduced complexity in requirements management due to multiple actors’ collaboration through several organizational boundaries.</div></div><div><h3>Objective:</h3><div>The main contribution of this article is to improve the understanding of requirements management in SECO. We propose a conceptual model whose concepts, definitions, and relationships are grounded in the literature and the modern software industry’s practices.</div></div><div><h3>Methods:</h3><div>We applied Design Science to build the conceptual model and conducted a Delphi study with 22 experts to assess it. We performed two rounds and adjusted our model according to the experts’ judgment.</div></div><div><h3>Results:</h3><div>We reached a conceptual model comprising 43 concepts and their relationships that help to understand requirements management in SECO. Moreover, we provided a glossary with a definition of each concept. This conceptual model can help abstract the complexity of the requirements management in SECO.</div></div><div><h3>Conclusions:</h3><div>By organizing concepts and relationships in requirements management in SECO, this conceptual model makes it possible to expand the body of knowledge in the area and serves as a basis for new solutions to support requirements management in SECO.</div></div>","PeriodicalId":54983,"journal":{"name":"Information and Software Technology","volume":"180 ","pages":"Article 107672"},"PeriodicalIF":3.8,"publicationDate":"2025-01-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143305313","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Alleviating class imbalance in Feature Envy prediction: An oversampling technique based on code entity attributes","authors":"Jiamin Guo ,&nbsp;Yangyang Zhao ,&nbsp;Tao Zheng ,&nbsp;Zhifei Chen ,&nbsp;Mingyue Jiang ,&nbsp;Zuohua Ding","doi":"10.1016/j.infsof.2025.107673","DOIUrl":"10.1016/j.infsof.2025.107673","url":null,"abstract":"<div><h3>Context:</h3><div>Feature Envy is a common code smell that occurs when a method heavily relies on data or functionality from other classes. Detecting Feature Envy is essential for improving software modularity and reducing technical debt. However, real-world datasets often exhibit severe class imbalance, with far fewer Feature Envy instances than non-smelly ones, posing challenges for prediction models. Traditional oversampling techniques attempt to address this issue by relying solely on numerical vectors but often fail to capture the complex relationships between code entities, potentially deviating from the nature of Feature Envy.</div></div><div><h3>Objective:</h3><div>This study introduces STANDER, a novel oversampling technique based on code entity similarity, designed to handle class imbalance in Feature Envy prediction by generating synthetic samples that better reflect the characteristics of Feature Envy.</div></div><div><h3>Method:</h3><div>STANDER creates synthetic samples by leveraging multidimensional code entity similarity, incorporating attributes such as dependency relationships, historical changes and code text. It was evaluated on five datasets using five classifiers: Naive Bayes, Logistic Regression, Support Vector Machine, Random Forest, and Decision Tree. Its performance was compared to baseline over-sampling techniques based on precision, recall, F1-score, and Matthews Correlation Coefficient.</div></div><div><h3>Results:</h3><div>STANDER enhances dataset diversity while maintaining clear boundaries between minority and majority classes, as reflected by higher Nearest Neighbor Diversity and Silhouette Score values. Models balanced with STANDER exhibited significant improvements in predictive performance, particularly in recall, F1-score, and Matthews Correlation Coefficient. Compared to the other oversampling techniques, STANDER demonstrated advantages in handling imbalanced datasets, especially in the Logistic Regression and Decision Tree classifiers. Statistical results confirm significant performance improvements across most models, highlighting its effectiveness and applicability.</div></div><div><h3>Conclusion:</h3><div>STANDER is an effective solution to alleviate class imbalance problem in Feature Envy detection by generating more representative synthetic samples that improve prediction performance.</div></div>","PeriodicalId":54983,"journal":{"name":"Information and Software Technology","volume":"180 ","pages":"Article 107673"},"PeriodicalIF":3.8,"publicationDate":"2025-01-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143305100","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"MPCA: Constructing the APTs provenance graphs through multi-perspective confidence and association","authors":"Zhao Zhang ,&nbsp;Senlin Luo ,&nbsp;Yingdan Guan ,&nbsp;Limin Pan","doi":"10.1016/j.infsof.2025.107670","DOIUrl":"10.1016/j.infsof.2025.107670","url":null,"abstract":"<div><div>The forensic analysis of Advanced Persistent Threats (APTs) attacks is crucial for maintaining cybersecurity. To address the challenges posed by the high complexity and strong concealment of APT attacks, provenance graph based on inter entity dependencies are used for forensic investigation. However, under long-term persistent attacks, entities with semantically consistent behavior patterns become excessively redundant, leading to an explosion of inter entity dependencies and a decrease in forensic efficiency. In addition, the implicit relationships within and between events are not fully represented, and alarm information spreads to neighboring benign events, making it difficult to accurately reconstruct attack scenario. In this paper, we propose an APT attack attribution method MPCA that combines multi-perspective confidence and association. Firstly, by merging parallel branches with semantically consistent behavior patterns in the process connected subgraph, redundant entities and their dependencies are reduced. Secondly, event confidence is estimated to exclude benign events, the association between events and alarms is analyzed to highlight attack events. Experimental results demonstrate that MPCA achieves state-of-the-art performance. MPCA can improve the efficiency of constructing attack scenario graphs, reduce false positive and false negative rates, and demonstrate greater adaptability in attack attribution tasks.</div></div>","PeriodicalId":54983,"journal":{"name":"Information and Software Technology","volume":"180 ","pages":"Article 107670"},"PeriodicalIF":3.8,"publicationDate":"2025-01-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143305099","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An input-denoising-based defense against stealthy backdoor attacks in large language models for code","authors":"Yubin Qu ,&nbsp;Song Huang ,&nbsp;Xiang Chen ,&nbsp;Tongtong Bai ,&nbsp;Yongming Yao","doi":"10.1016/j.infsof.2024.107661","DOIUrl":"10.1016/j.infsof.2024.107661","url":null,"abstract":"<div><h3>Context:</h3><div>Large Language Models are becoming integral to software development. They are trained on open data from platforms like GitHub, making them vulnerable to poisoning attacks. Research shows that backdoor attacks with traditional static triggers using fixed code patterns are relatively easy to detect. The novel attack approach uses specific Syntax Tree structures as triggers, offering greater stealthiness while maintaining explicit code structures. This method poses new challenges for backdoor detection.</div></div><div><h3>Objective:</h3><div>We propose an <strong>I</strong>nput-<strong>D</strong> enoising-based defense against stealthy <strong>B</strong>ackdoor <strong>A</strong>ttacks with dynamic triggers (<em>IDBA</em>) in Large Language Models for Code.</div></div><div><h3>Method:</h3><div>We overlay a set of malicious code segments onto the code segment with dynamic triggers, convert the output state of the input code into a random walk graph neural network, calculate the expected value of the final state through particle filtering, and thus detect the existence of a backdoor attack.</div></div><div><h3>Results:</h3><div>Empirical studies are conducted on Codebert, GraphCodebert, and CodeT5 for vulnerability and code clone detection tasks. Our results show that <em>IDBA</em> achieves an average detection rate of 73.75% and 68.12% for vulnerability and code clone detection tasks, respectively.</div></div><div><h3>Conclusion:</h3><div>Detecting backdoor attacks using <em>IDBA</em> on code models allows for the early identification of potential backdoor threats after model deployment, enhancing the security of code models.</div></div>","PeriodicalId":54983,"journal":{"name":"Information and Software Technology","volume":"180 ","pages":"Article 107661"},"PeriodicalIF":3.8,"publicationDate":"2025-01-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143305098","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"MT-Nod: Metamorphic testing for detecting non-optimal decisions of autonomous driving systems in interactive scenarios","authors":"Zhen Yang ,&nbsp;Song Huang ,&nbsp;Xingya Wang ,&nbsp;Tongtong Bai ,&nbsp;Yang Wang","doi":"10.1016/j.infsof.2024.107659","DOIUrl":"10.1016/j.infsof.2024.107659","url":null,"abstract":"<div><h3>Context:</h3><div>Autonomous driving technology advances into daily life, with expectations for autonomous driving systems (ADSs) to make optimal, human-like decisions. However, ADSs often exhibit “unintelligent” behaviors like inefficient path choices, significantly impacting travel efficiency and potentially causing delays. Therefore, testing the decision optimality of ADSs is critically urgent. However, the testing process faces a significant “testing oracle” problem, and current methods overlook behavior interactions, which do not reflect real-world traffic scenarios.</div></div><div><h3>Objective:</h3><div>To assess the performance and reliability of ADSs in optimal decision-making, mitigate test oracle problems, and detect non-optimal decisions without calculating the optimal path.</div></div><div><h3>Method:</h3><div>This paper proposes a metamorphic testing method for optimal decision-making in autonomous driving under interactive scenarios, MT-Nod. Our method introduces a novel metamorphic relation to evaluate the optimality of path directions, along with a follow-up scenario generation method. The scenario generation method includes mutation points selection based on key behaviors, scenario mutation based on behavioral interactions, and road accessibility assessment, to generate scenarios with dynamic interactions. Additionally, a scenario scheduling strategy is designed to prioritize and schedule scenarios based on the priority of mutation points.</div></div><div><h3>Results:</h3><div>We evaluate MT-Nod extensively on the advanced Apollo ADS. Across four source scenarios, it generates 69.3 non-optimal decision scenarios (NoDSs), classified into eight types. Compared to baselines, MT-Nod efficiently produces and detects a greater variety and quantity of NoDSs.</div></div><div><h3>Conclusion:</h3><div>The proposed method for optimal decision testing under interactive scenarios, MT-Nod, effectively detects non-optimal decisions of ADSs. These “unintelligent” behaviors are crucial for enhancing the performance and reliability of ADSs.</div></div>","PeriodicalId":54983,"journal":{"name":"Information and Software Technology","volume":"180 ","pages":"Article 107659"},"PeriodicalIF":3.8,"publicationDate":"2025-01-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143304461","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Boosting mutation-based fault localization by effectively generating Higher-Order Mutants","authors":"Shumei Wu ,&nbsp;Binbin Yang ,&nbsp;Zexing Chang ,&nbsp;Zheng Li ,&nbsp;Xiang Chen ,&nbsp;Yong Liu","doi":"10.1016/j.infsof.2024.107660","DOIUrl":"10.1016/j.infsof.2024.107660","url":null,"abstract":"&lt;div&gt;&lt;h3&gt;Context:&lt;/h3&gt;&lt;div&gt;Fault Localization (FL) is an important and tedious phase of software debugging. Among various FL techniques, Mutation-Based Fault Localization (MBFL) demonstrates promising FL accuracy utilizing impact information of statements provided by First-Order-Mutants (FOMs). Despite its success in Single-Fault Scenarios (SFSs), it fails to achieve satisfactory performance in Multiple-Fault Scenarios (MFSs). Higher-Order-Mutants (HOMs) provide a solution for MFSs. However, existing work on HOM generation is inadequate and ignores the correlation among faults in MFSs.&lt;/div&gt;&lt;/div&gt;&lt;div&gt;&lt;h3&gt;Objective:&lt;/h3&gt;&lt;div&gt;In this article, we systematically analyze three relationships among single-faults in MFSs, and further propose three HOM generation methods (i.e., SFClu, SFDis, and SFDen) to simulate different multiple-faults and improve the effectiveness of MBFL in MFSs.&lt;/div&gt;&lt;/div&gt;&lt;div&gt;&lt;h3&gt;Method:&lt;/h3&gt;&lt;div&gt;We investigate the multiple-fault composition on real-world 393 faulty programs from Defects4J, and then apply our methods to generate HOMs for FL. Specifically, SFClu focuses on generating appropriate HOMs for Multi-Single-Source Fault (MSSF) scenarios, where each single-fault is responsible for different observed failures. SFDis is well-suited for Multi-Coupled-Source Fault (MCSF) scenarios where at least two single-faults can interact with each other, leading to certain failures either being observable or masked. SFDen aims to generate suitable HOMs for Single-Coupled-Source Fault (SCSF) scenarios with multiple single-faults that occur within a statement.&lt;/div&gt;&lt;/div&gt;&lt;div&gt;&lt;h3&gt;Results:&lt;/h3&gt;&lt;div&gt;(1) The proportion of MFSs is as high as 63.10% in real-world programs, with MSSF, MCSF , and SCSF scenarios accounting for 35.08%, 53.23%, and 11.69%, respectively. (2) Compared to the best-performing mutant generation method Neural-MBFL, SFClu, SFDis, and SFDen can improve the FL performance for MBFL by 36.78%, 49.80%, and 16.36% in &lt;span&gt;&lt;math&gt;&lt;mrow&gt;&lt;mi&gt;T&lt;/mi&gt;&lt;mi&gt;o&lt;/mi&gt;&lt;mi&gt;p&lt;/mi&gt;&lt;/mrow&gt;&lt;/math&gt;&lt;/span&gt;-1, respectively, outperforming eight established SBFL and MBFL techniques. (3) SFClu, SFDis, and SFDen are more suitable for MSSF, MCSF, and SCSF scenarios, respectively, which aligns with their design intend. (4) Their combination further enhances FL accuracy, achieving up to 85 faults successfully localized and an average improvement of 29.54% in &lt;span&gt;&lt;math&gt;&lt;mrow&gt;&lt;mi&gt;T&lt;/mi&gt;&lt;mi&gt;o&lt;/mi&gt;&lt;mi&gt;p&lt;/mi&gt;&lt;/mrow&gt;&lt;/math&gt;&lt;/span&gt;-1. Finally, extensive evaluations on SIR with artificial faults and Codeflaws containing student programs demonstrate the generalization of SFClu, the applicability of SFDis on real faults, and the suitability of SFDen for student programs.&lt;/div&gt;&lt;/div&gt;&lt;div&gt;&lt;h3&gt;Conclusion:&lt;/h3&gt;&lt;div&gt;Empirical studies have confirmed the prevalence of MFSs, highlighting the significance of multi-fault localization. Moreover, our proposed three HOM generation methods can further enhance the performance of MBFL with HOMs in MFSs, showing their effectivenes","PeriodicalId":54983,"journal":{"name":"Information and Software Technology","volume":"180 ","pages":"Article 107660"},"PeriodicalIF":3.8,"publicationDate":"2025-01-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143305097","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Concept definition review: A method for studying terminology in software engineering","authors":"Sabine Molenaar,&nbsp;Nikita van den Berg,&nbsp;Fabiano Dalpiaz,&nbsp;Sjaak Brinkkemper","doi":"10.1016/j.infsof.2024.107648","DOIUrl":"10.1016/j.infsof.2024.107648","url":null,"abstract":"<div><h3>Context:</h3><div>In scientific domains, definitions provide a precise description of fundamental concepts. Although the debate within the philosophy of computer science regarding the scientific nature of software engineering (SE) is inconclusive, SE researchers have laid down important steps toward treating SE as a scientific paradigm.</div></div><div><h3>Objective:</h3><div>We aim to support precise and effective communication among SE researchers and practitioners by providing a systematic process for the identification and analysis of definitions, in order to support the selection of a suitable definition for a certain use case.</div></div><div><h3>Method:</h3><div>Inspired by methods for the planning and execution of systematic literature reviews, we construct a method that is specific for concept definition reviews (CDRs). These reviews are performed whenever a research team wishes to obtain a detailed understanding of an SE concept that may have been characterized by dozens, if not hundreds, definitions.</div></div><div><h3>Results:</h3><div>We built our method via two <em>design science</em> iterations. The first one focused on the concept <em>feature</em> and resulted in the definitive version of the CDR method presented in this paper. We then applied the revised method to two, related concepts: <em>quality requirement</em> and <em>non-functional requirement</em>. Besides showing the applicability of the CDR method, our results include findings regarding the characteristics and evolution of the terms.</div></div><div><h3>Conclusions:</h3><div>The two applications of the CDR method highlight the existence and citation of hundreds of definitions, many of which are nearly (but not exactly) identical. We put forward our method for other researchers to shed light on the key terminology in other sub-fields of SE.</div></div>","PeriodicalId":54983,"journal":{"name":"Information and Software Technology","volume":"180 ","pages":"Article 107648"},"PeriodicalIF":3.8,"publicationDate":"2024-12-31","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143304456","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Data analytics in software startups: Understanding key concepts and critical challenges","authors":"Usman Rafiq,&nbsp;Xiaofeng Wang,&nbsp;Eduardo Guerra","doi":"10.1016/j.infsof.2024.107652","DOIUrl":"10.1016/j.infsof.2024.107652","url":null,"abstract":"<div><h3>Context:</h3><div>The continuous proliferation of data nowadays has inspired companies to make data-informed decisions. Despite the acknowledged benefits of analytics, there is a persistent question about how companies, especially software startup companies with distinguishing characteristics, can effectively create value from it. In the startup context, analytics refers to the use of startup data and insights to inform strategies and tactics across startup business, product, team, sales, and marketing dimensions.</div></div><div><h3>Objective:</h3><div>In this study, we aim to bridge the knowledge gap by eliciting an understanding of the analytics that software startup companies hold and identifying critical challenges they face in the realm of analytics.</div></div><div><h3>Method:</h3><div>We conducted a multiple-case study with eight software startups at different startup stages. In addition to the data collected through semi-structured interviews, we considered other data sources such as analytics dashboards and online data about the startups, including websites and social media platforms. We analyzed the data using thematic analysis.</div></div><div><h3>Results:</h3><div>Our results firstly revealed a divergent understanding of analytics by software startups, based on which we reported essential characteristics of analytics perceived by them. Then we identified 22 analytics challenges classified into six main themes. The themes encompass data capture and access challenges, data interpretation and bias, communication challenges, cultural challenges, external influences and constraints, and analytics implementation challenges.</div></div><div><h3>Conclusions:</h3><div>Our findings contribute to a conceptual understanding of analytics in software startups and the identification of critical challenges faced by these startups across different stages. The conceptual understanding lays the foundation for comprehending what constitutes analytics for software startups, while the identification of challenges anticipates critical barriers to the adoption and implementation of analytics. We also provide practical implications to both researchers and practitioners.</div></div>","PeriodicalId":54983,"journal":{"name":"Information and Software Technology","volume":"180 ","pages":"Article 107652"},"PeriodicalIF":3.8,"publicationDate":"2024-12-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143305312","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Introduction for Special Issue on Software Production","authors":"","doi":"10.1016/j.infsof.2024.107658","DOIUrl":"10.1016/j.infsof.2024.107658","url":null,"abstract":"","PeriodicalId":54983,"journal":{"name":"Information and Software Technology","volume":"179 ","pages":"Article 107658"},"PeriodicalIF":3.8,"publicationDate":"2024-12-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143128052","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"What are the emotions of developers towards deep learning documentation? — An exploratory study on Stack Overflow posts","authors":"Akhila Sri Manasa Venigalla,&nbsp;Sridhar Chimalakonda","doi":"10.1016/j.infsof.2024.107655","DOIUrl":"10.1016/j.infsof.2024.107655","url":null,"abstract":"<div><h3>Context:</h3><div>Non native machine learning and deep learning (DL) developers face several challenges in using DL frameworks owing to the issues persistent in DL documentation. However, there are no studies that explore the reasons for issues in documentation.</div></div><div><h3>Objective:</h3><div>Investigating the underlying emotions in developer discussions on documentation could help in identifying reasons for issues in documentation. Hence, in this study, we analyse emotions of Stack Overflow posts corresponding to documentation of DL frameworks.</div></div><div><h3>Methodology:</h3><div>We identify relevant deep-learning related tags using integrated snowballing approach and extract 159.2K posts related to DL. We then identify documentation related posts among these using keyword matching approach, which resulted in 13,572 DL documentation related posts. We use Random Forest Classifier to build six emotion classifier models based on Gold Label Dataset for emotions. We then classify the extracted posts into each of the six emotions — <em>Anger</em>, <em>Fear</em>, <em>Love</em>, <em>Joy</em>, <em>Sadness</em> and <em>Surprise</em> using the classifier models, and curate the results.</div></div><div><h3>Results:</h3><div>We observe a large expression of anger and sadness, with more than half of posts having ‘yolo’ and ‘activation-function’ tags exhibiting these emotions, while <em>Love</em> emotion is predominantly present in posts with ‘theano’ tag. During our analysis, we observed that 40% of ‘Body’ and ‘Answer’ posts exhibited anger and sadness emotions.</div></div><div><h3>Conclusion:</h3><div>Our study reveals the large presence of Anger, Fear and Sadness emphasizing the need to improve DL framework documentation. Specifically, maintainers of the ‘yolo’ and ‘matcaffe’ libraries could improve their documentation, as the corresponding posts exhibit more of <em>Anger</em> and <em>Sadness</em>.</div></div>","PeriodicalId":54983,"journal":{"name":"Information and Software Technology","volume":"179 ","pages":"Article 107655"},"PeriodicalIF":3.8,"publicationDate":"2024-12-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143091907","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}