GBM Returns the Best Prediction Performance among Regression Approaches: A Case Study of Stack Overflow Code Quality

• URL: http://arxiv.org/abs/2505.10019v1
• Date: Thu, 15 May 2025 07:04:17 GMT
• Title: GBM Returns the Best Prediction Performance among Regression Approaches: A Case Study of Stack Overflow Code Quality
• Authors: Sherlock A. Licorish, Brendon Woodford, Lakmal Kiyaduwa Vithanage, Osayande Pascal Omondiagbe,
• Abstract summary: We examined the variables that predict Stack Overflow (Java) code quality, and the regression approach that provides the best predictive power.<n>Longer Stack Overflow code tended to have more code violations, questions that were scored higher also attracted more views and the more answers that are added to questions on Stack Overflow the more errors were typically observed in the code that was provided.
• Score: 2.5515299924109858
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Practitioners are increasingly dependent on publicly available resources for supporting their knowledge needs during software development. This has thus caused a spotlight to be paced on these resources, where researchers have reported mixed outcomes around the quality of these resources. Stack Overflow, in particular, has been studied extensively, with evidence showing that code resources on this platform can be of poor quality at times. Limited research has explored the variables or factors that predict code quality on Stack Overflow, but instead has focused on ranking content, identifying defects and predicting future content. In many instances approaches used for prediction are not evaluated to identify the best techniques. Contextualizing the Stack Overflow code quality problem as regression-based, we examined the variables that predict Stack Overflow (Java) code quality, and the regression approach that provides the best predictive power. Six approaches were considered in our evaluation, where Gradient Boosting Machine (GBM) stood out. In addition, longer Stack Overflow code tended to have more code violations, questions that were scored higher also attracted more views and the more answers that are added to questions on Stack Overflow the more errors were typically observed in the code that was provided. Outcomes here point to the value of the GBM ensemble learning mechanism, and the need for the practitioner community to be prudent when contributing and reusing Stack Overflow Java coding resource.

Related papers

• Do AI models help produce verified bug fixes? [62.985237003585674]
  Large Language Models are used to produce corrections to software bugs.<n>This paper investigates how programmers use Large Language Models to complement their own skills.<n>The results are a first step towards a proper role for AI and LLMs in providing guaranteed-correct fixes to program bugs.
  arXiv  Detail & Related papers  (2025-07-21T17:30:16Z)
• Understanding Code Understandability Improvements in Code Reviews [79.16476505761582]
  We analyzed 2,401 code review comments from Java open-source projects on GitHub. 83.9% of suggestions for improvement were accepted and integrated, with fewer than 1% later reverted.
  arXiv  Detail & Related papers  (2024-10-29T12:21:23Z)
• Towards Better Answers: Automated Stack Overflow Post Updating [11.85319691188159]
  We introduce a novel framework, named Soup (Stack Overflow Updator for Post) for this task. Soup addresses two key tasks: Valid Comment-Edit Prediction (VCP) and Automatic Post Updating (APU)
  arXiv  Detail & Related papers  (2024-08-17T04:48:53Z)
• Investigating the Transferability of Code Repair for Low-Resource Programming Languages [57.62712191540067]
  Large language models (LLMs) have shown remarkable performance on code generation tasks. Recent works augment the code repair process by integrating modern techniques such as chain-of-thought reasoning or distillation. We investigate the benefits of distilling code repair for both high and low resource languages.
  arXiv  Detail & Related papers  (2024-06-21T05:05:39Z)
• CodeRAG-Bench: Can Retrieval Augment Code Generation? [78.37076502395699]
  We conduct a systematic, large-scale analysis of code generation using retrieval-augmented generation.<n>We first curate a comprehensive evaluation benchmark, CodeRAG-Bench, encompassing three categories of code generation tasks.<n>We examine top-performing models on CodeRAG-Bench by providing contexts retrieved from one or multiple sources.
  arXiv  Detail & Related papers  (2024-06-20T16:59:52Z)
• A Tale of Two Communities: Exploring Academic References on Stack Overflow [1.2914230269240388]
  We find that Stack Overflow communities with different domains of interest engage with academic literature at varying frequencies and speeds. The contradicting patterns suggest that some disciplines may have diverged in their interests and development trajectories from the corresponding practitioner community.
  arXiv  Detail & Related papers  (2024-03-14T20:33:55Z)
• Answer ranking in Community Question Answering: a deep learning approach [0.0]
  This work tries to advance the state of the art on answer ranking for community Question Answering by proceeding with a deep learning approach. We created a large data set of questions and answers posted to the Stack Overflow website. We leveraged the natural language processing capabilities of dense embeddings and LSTM networks to produce a prediction for the accepted answer attribute.
  arXiv  Detail & Related papers  (2022-10-16T18:47:41Z)
• Features that Predict the Acceptability of Java and JavaScript Answers on Stack Overflow [5.332217496693262]
  We studied the Stack Overflow dataset by analyzing questions and answers for the two most popular tags (Java and JavaScript) Our findings reveal that the length of code in answers, reputation of users, similarity of the text between questions and answers, and the time lag between questions and answers have the highest predictive power for differentiating accepted and unaccepted answers.
  arXiv  Detail & Related papers  (2021-01-08T03:09:38Z)
• The Influence of Domain-Based Preprocessing on Subject-Specific Clustering [55.41644538483948]
  The sudden change of moving the majority of teaching online at Universities has caused an increased amount of workload for academics. One way to deal with this problem is to cluster these questions depending on their topic. In this paper, we explore the realms of tagging data sets, focusing on identifying code excerpts and providing empirical results.
  arXiv  Detail & Related papers  (2020-11-16T17:47:19Z)
• Retrieve, Program, Repeat: Complex Knowledge Base Question Answering via Alternate Meta-learning [56.771557756836906]
  We present a novel method that automatically learns a retrieval model alternately with the programmer from weak supervision. Our system leads to state-of-the-art performance on a large-scale task for complex question answering over knowledge bases.
  arXiv  Detail & Related papers  (2020-10-29T18:28:16Z)
• StackOverflow vs Kaggle: A Study of Developer Discussions About Data Science [0.0]
  This paper conducts experiments to study the characteristics of 197836 posts from StackOverflow and Kaggle. The main findings include that-related topics were most prevalent in StackOverflow. Across both communities, DS discussion is increasing at a dramatic rate. ensemble algorithms are the most mentioned ML/DL algorithms in Kaggle but are rarely discussed on StackOverflow.
  arXiv  Detail & Related papers  (2020-06-06T06:51:11Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.