Towards an Understanding of Context Utilization in Code Intelligence

Code intelligence is an emerging domain in software engineering, aiming to improve the effectiveness and efficiency of various code-related tasks. Recent research suggests that incorporating contextual information beyond the basic original task inputs (i.e., source code) can substantially enhance model performance. Such contextual signals may be obtained directly or indirectly from sources such as API documentation or intermediate representations like abstract syntax trees can significantly improve the effectiveness of code intelligence. Despite growing academic interest, there is a lack of systematic analysis of context in code intelligence. To address this gap, we conduct an extensive literature review of 146 relevant studies published between September 2007 and August 2024. Our investigation yields four main contributions. (1) A quantitative analysis of the research landscape, including publication trends, venues, and the explored domains; (2) A novel taxonomy of context types used in code intelligence; (3) A task-oriented analysis investigating context integration strategies across diverse code intelligence tasks; (4) A critical evaluation of evaluation methodologies for context-aware methods. Based on these findings, we identify fundamental challenges in context utilization in current code intelligence systems and propose a research roadmap that outlines key opportunities for future research.

How Well Do LLMs Generate Code for Different Application Domains? Benchmark and Evaluation

Recently, an increasing number of AI-driven programming assistants powered by code LLMs have been integrated into various real-world software development environments, significantly boosting developer productivity. However, existing code generation benchmarks primarily focus on general-purpose scenarios, leaving the code generation performance of LLMs for specific application domains largely unknown. In this paper, we introduce a new benchmark, MultiCodeBench, to fill this gap. MultiCodeBench comprises 2,400 programming tasks, covering 12 popular software development domains and 15 programming languages. Specifically, we perform in-depth research to identify these 12 application domains. Given that each domain may involve multiple technical frameworks, and that different frameworks present distinct challenges in the coding process, we categorize the commonly used frameworks and platforms within each domain. We then sample programming problems from GitHub repositories related to these subdomains. To ensure the quality of the tasks and mitigate data leakage issues, we invite annotators to rewrite the docstrings for each task in MultiCodeBench. Additionally, we build a static analysis-based dependency parsing tool to extract the dependencies in the ground truth for each task, enabling deeper performance analysis. Through extensive experiments on MultiCodeBench with eleven representative mainstream LLMs, we reveal the code generation performance of the LLMs across different application domains, providing practical insights for developers in downstream fields when selecting LLMs. Furthermore, we analyze the reasons behind the models' failures in completing software application development tasks, offering guidance for model developers to enhance domain-specific code generation capabilities.

Agents in Software Engineering: Survey, Landscape, and Vision

In recent years, Large Language Models (LLMs) have achieved remarkable success and have been widely used in various downstream tasks, especially in the tasks of the software engineering (SE) field. We find that many studies combining LLMs with SE have employed the concept of agents either explicitly or implicitly. However, there is a lack of an in-depth survey to sort out the development context of existing works, analyze how existing works combine the LLM-based agent technologies to optimize various tasks, and clarify the framework of LLM-based agents in SE. In this paper, we conduct the first survey of the studies on combining LLM-based agents with SE and present a framework of LLM-based agents in SE which includes three key modules: perception, memory, and action. We also summarize the current challenges in combining the two fields and propose future opportunities in response to existing challenges. We maintain a GitHub repository of the related papers at: https://github.com/DeepSoftwareAnalytics/Awesome-Agent4SE.

SoTaNa: The Open-Source Software Development Assistant

Software development plays a crucial role in driving innovation and efficiency across modern societies. To meet the demands of this dynamic field, there is a growing need for an effective software development assistant. However, existing large language models represented by ChatGPT suffer from limited accessibility, including training data and model weights. Although other large open-source models like LLaMA have shown promise, they still struggle with understanding human intent. In this paper, we present SoTaNa, an open-source software development assistant. SoTaNa utilizes ChatGPT to generate high-quality instruction-based data for the domain of software engineering and employs a parameter-efficient fine-tuning approach to enhance the open-source foundation model, LLaMA. We evaluate the effectiveness of \our{} in answering Stack Overflow questions and demonstrate its capabilities. Additionally, we discuss its capabilities in code summarization and generation, as well as the impact of varying the volume of generated data on model performance. Notably, SoTaNa can run on a single GPU, making it accessible to a broader range of researchers. Our code, model weights, and data are public at \url{https://github.com/DeepSoftwareAnalytics/SoTaNa}.

Towards Efficient Fine-tuning of Pre-trained Code Models: An Experimental Study and Beyond

Recently, fine-tuning pre-trained code models such as CodeBERT on downstream tasks has achieved great success in many software testing and analysis tasks. While effective and prevalent, fine-tuning the pre-trained parameters incurs a large computational cost. In this paper, we conduct an extensive experimental study to explore what happens to layer-wise pre-trained representations and their encoded code knowledge during fine-tuning. We then propose efficient alternatives to fine-tune the large pre-trained code model based on the above findings. Our experimental study shows that (1) lexical, syntactic and structural properties of source code are encoded in the lower, intermediate, and higher layers, respectively, while the semantic property spans across the entire model. (2) The process of fine-tuning preserves most of the code properties. Specifically, the basic code properties captured by lower and intermediate layers are still preserved during fine-tuning. Furthermore, we find that only the representations of the top two layers change most during fine-tuning for various downstream tasks. (3) Based on the above findings, we propose Telly to efficiently fine-tune pre-trained code models via layer freezing. The extensive experimental results on five various downstream tasks demonstrate that training parameters and the corresponding time cost are greatly reduced, while performances are similar or better. Replication package including source code, datasets, and online Appendix is available at: \url{https://github.com/DeepSoftwareAnalytics/Telly}.

Enhancing Semantic Code Search with Multimodal Contrastive Learning and Soft Data Augmentation

Code search aims to retrieve the most semantically relevant code snippet for a given natural language query. Recently, large-scale code pre-trained models such as CodeBERT and GraphCodeBERT learn generic representations of source code and have achieved substantial improvement on code search task. However, the high-quality sequence-level representations of code snippets have not been sufficiently explored. In this paper, we propose a new approach with multimodal contrastive learning and soft data augmentation for code search. Multimodal contrastive learning is used to pull together the representations of code-query pairs and push apart the unpaired code snippets and queries. Moreover, data augmentation is critical in contrastive learning for learning high-quality representations. However, only semantic-preserving augmentations for source code are considered in existing work. In this work, we propose to do soft data augmentation by dynamically masking and replacing some tokens in code sequences to generate code snippets that are similar but not necessarily semantic-preserving as positive samples for paired queries. We conduct extensive experiments to evaluate the effectiveness of our approach on a large-scale dataset with six programming languages. The experimental results show that our approach significantly outperforms the state-of-the-art methods. We also adapt our techniques to several pre-trained models such as RoBERTa and CodeBERT, and significantly boost their performance on the code search task.

On the Evaluation of Commit Message Generation Models: An Experimental Study

Commit messages are natural language descriptions of code changes, which are important for program understanding and maintenance. However, writing commit messages manually is time-consuming and laborious, especially when the code is updated frequently. Various approaches utilizing generation or retrieval techniques have been proposed to automatically generate commit messages. To achieve a better understanding of how the existing approaches perform in solving this problem, this paper conducts a systematic and in-depth analysis of the state-of-the-art models and datasets. We find that: (1) Different variants of the BLEU metric are used in previous works, which affects the evaluation and understanding of existing methods. (2) Most existing datasets are crawled only from Java repositories while repositories in other programming languages are not sufficiently explored. (3) Dataset splitting strategies can influence the performance of existing models by a large margin. Some models show better performance when the datasets are split by commit, while other models perform better when the datasets are split by timestamp or by project. Based on our findings, we conduct a human evaluation and find the BLEU metric that best correlates with the human scores for the task. We also collect a large-scale, information-rich, and multi-language commit message dataset MCMD and evaluate existing models on this dataset. Furthermore, we conduct extensive experiments under different dataset splitting strategies and suggest the suitable models under different scenarios. Based on the experimental results and findings, we provide feasible suggestions for comprehensively evaluating commit message generation models and discuss possible future research directions. We believe this work can help practitioners and researchers better evaluate and select models for automatic commit message generation.

Neural Code Summarization: How Far Are We?

Source code summaries are important for the comprehension and maintenance of programs. However, there are plenty of programs with missing, outdated, or mismatched summaries. Recently, deep learning techniques have been exploited to automatically generate summaries for given code snippets. To achieve a profound understanding of how far we are from solving this problem, in this paper, we conduct a systematic and in-depth analysis of five state-of-the-art neural source code summarization models on three widely used datasets. Our evaluation results suggest that: (1) The BLEU metric, which is widely used by existing work for evaluating the performance of the summarization models, has many variants. Ignoring the differences among the BLEU variants could affect the validity of the claimed results. Furthermore, we discover an important, previously unknown bug about BLEU calculation in a commonly-used software package. (2) Code pre-processing choices can have a large impact on the summarization performance, therefore they should not be ignored. (3) Some important characteristics of datasets (corpus size, data splitting method, and duplication ratio) have a significant impact on model evaluation. Based on the experimental results, we give some actionable guidelines on more systematic ways for evaluating code summarization and choosing the best method in different scenarios. We also suggest possible future research directions. We believe that our results can be of great help for practitioners and researchers in this interesting area.

Is a Single Model Enough? MuCoS: A Multi-Model Ensemble Learning for Semantic Code Search

Recently, deep learning methods have become mainstream in code search since they do better at capturing semantic correlations between code snippets and search queries and have promising performance. However, code snippets have diverse information from different dimensions, such as business logic, specific algorithm, and hardware communication, so it is hard for a single code representation module to cover all the perspectives. On the other hand, as a specific query may focus on one or several perspectives, it is difficult for a single query representation module to represent different user intents. In this paper, we propose MuCoS, a multi-model ensemble learning architecture for semantic code search. It combines several individual learners, each of which emphasizes a specific perspective of code snippets. We train the individual learners on different datasets which contain different perspectives of code information, and we use a data augmentation strategy to get these different datasets. Then we ensemble the learners to capture comprehensive features of code snippets.