Search arXiv⌕ Search
Home/Archive/arXiv

arXiv · 2310.02104

An empirical study of ChatGPT-3.5 on question answering and code maintenance

Md Mahir Asef Kabir·Sk Adnan Hassan·Xiaoyin Wang·Ying Wang·Hai Yu·Na Meng
View original source ↗Read source document ↗Export metadata ↓

Abstract

Ever since the launch of ChatGPT in 2022, a rising concern is whether ChatGPT will replace programmers and kill jobs. Motivated by this widespread concern, we conducted an empirical study to systematically compare ChatGPT against programmers in question-answering and software-maintaining. We reused a dataset introduced by prior work, which includes 130 StackOverflow (SO) discussion threads referred to by the Java developers of 357 GitHub projects. We mainly investigated three research questions (RQs). First, how does ChatGPT compare with programmers when answering technical questions? Second, how do developers perceive the differences between ChatGPT's answers and SO answers? Third, how does ChatGPT compare with humans when revising code for maintenance requests? For RQ1, we provided the 130 SO questions to ChatGPT, and manually compared ChatGPT answers with the accepted/most popular SO answers in terms of relevance, readability, informativeness, comprehensiveness, and reusability. For RQ2, we conducted a user study with 30 developers, asking each developer to assess and compare 10 pairs of answers, without knowing the information source (i.e., ChatGPT or SO). For RQ3, we distilled 48 software maintenance tasks from 48 GitHub projects citing the studied SO threads. We queried ChatGPT to revise a given Java file, and to incorporate the code implementation for any prescribed maintenance requirement. Our study reveals interesting phenomena: For the majority of SO questions (97/130), ChatGPT provided better answers; in 203 of 300 ratings, developers preferred ChatGPT answers to SO answers; ChatGPT revised code correctly for 22 of the 48 tasks. Our research will expand people's knowledge of ChatGPT capabilities, and shed light on future adoption of ChatGPT by the software industry.

Explore related subjects

cs.SE

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Md Mahir Asef Kabir, Sk Adnan Hassan, Xiaoyin Wang, Ying Wang, Hai Yu, Na Meng. 2023-10-03. An empirical study of ChatGPT-3.5 on question answering and code maintenance. https://arxiv.org/abs/2310.02104

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

XScientist: A Git-Like Research Protocol for Long-Running Autonomous Scientific Discovery

Autonomous research systems can generate plausible papers while losing the decisions, failed branches, and evidence needed to inspect or continue the work. We present XScientist, a local-first, git-like protocol that treats research state, rather than a manuscript, as the unit of continuation. Hypotheses, experiment attempts, observations, claims, reviews, and handoffs are represented as typed, content-addressed objects in an exploration graph. Immutable checkpoints, explicit negative outcomes, claim-evidence closure, replay boundaries, and authority-aware gates make each transition inspectable without treating a passing integrity check as scientific truth. The protocol exports a portable Agent-Native Research Artifact (ARA) that another agent or human can inspect, fork, verify, and extend. A reference implementation integrates planning, execution, review, repair, and supervised long-running operation while preserving provenance across these stages. We evaluate the protocol with controlled artifact-integrity workloads and matched external task pilots, keeping native task performance separate from evidence and audit claims. The result is an interoperability and accountability layer for long-running autonomous science, with explicit boundaries where human judgment and independent evaluation remain necessary.

cs.SE↗
arXiv2026

AutoSQL: Extracting SQL Templates from Imperative ORM Code in Large-Scale Repositories

Suboptimal SQL queries can significantly degrade the performance of cloud systems, motivating the extraction and auditing of SQL statements before deployment. However, Go ORM frameworks construct SQL imperatively through scattered method-call sequences, making it difficult to statically recover the resulting SQL templates. We present AutoSQL, a system that reconstructs SQL templates from Go ORM code. AutoSQL constructs a Code Index, a directed graph that captures structural dependencies between functions, types, and global variables as navigable edges. It then traces upstream call chains from ORM invocation sites to identify database-interacting functions as entry points. For each entry point, an LLM agent traverses the Code Index to collect code slices that influence SQL generation, switching to pattern-based search when the graph cannot resolve a retrieval goal. We call this strategy Hybrid Context Retrieval. Once sufficient context is collected, the agent synthesizes SQL templates. Evaluation on a benchmark of 579 test-covered entry points and 1,186 runtime-traced SQL statements from five large-scale Go repositories shows that AutoSQL achieves 68.04% to 72.18% recall, exceeding the static reachability baseline by 11.80% to 15.94% and outperforming existing methods by 8.52% to 21.50%.

cs.SE↗
arXiv2026

The Vocabulary of Flaky Tests in Swift

Flaky tests produce non-deterministic outcomes without code change, eroding CI confidence and delaying deliveries. While vocabulary-based machine learning prediction has proven effective for Java and JavaScript, no study has evaluated it for Swift, a language whose testing style is dominated by UI and asynchronous code. We collect 91 flaky and 22,349 stable tests from 15 open-source Swift projects via re-execution and commit-history mining, then train five classifiers (Random Forest, Decision Tree, Naive Bayes, SVM, KNN) on TF-IDF unigram+bigram features under stratified 5-fold cross-validation. Random Forest achieves the best performance (Precision = 0.92, F1 = 0.86, AUC = 0.95) and substantially outperforms trivial baselines, among them a vocabulary-threshold rule applied to the most informative tokens, confirming a genuine discriminative signal (MCC = 0.75 vs. 0.08 for the best baseline). Information-gain analysis reveals two complementary signal types. Flakiness markers appear predominantly in unstable tests and comprise concurrency primitives (async, await), expectation-based synchronisation (expectation, fulfill), error propagation (throws), and explicit timing dependence (timeout, wait, now). Stability markers, chiefly the assertion vocabulary of plainly synchronous tests (xctassertequal), count as evidence against flakiness. Error analysis shows that the model fails when flakiness is hidden in shared infrastructure outside the test body or when async constructs are used in a deterministic context, exposing the intrinsic limit of lexical prediction. These results extend vocabulary-based flakiness detection to the Swift ecosystem and characterise both its effectiveness and its boundaries.

cs.SE↗