اشترك بالحزمة الذهبية واحصل على وصول غير محدود شمرا أكاديميا
تسجيل مستخدم جديدOn the Effectiveness of Transfer Learning for Code Search
127
0
0.0
(
0
)
اسأل ChatGPT حول البحث
ﻻ يوجد ملخص باللغة العربية
The Transformer architecture and transfer learning have marked a quantum leap in natural language processing, improving the state of the art across a range of text-based tasks. This paper examines how these advancements can be applied to and improve code search. To this end, we pre-train a BERT-based model on combinations of natural language and source code data and evaluate it on pairs of StackOverflow question titles and code answers. Our results show that the pre-trained models consistently outperform the models that were not pre-trained. In cases where the model was pre-trained on natural language and source code data, it also outperforms an information retrieval baseline based on Lucene. Also, we demonstrated that combined use of an information retrieval-based approach followed by a Transformer, leads to the best results overall, especially when searching into a large search pool. Furthermore, transfer learning is particularly effective when much pre-training data is available and fine-tuning data is limited. We demonstrate that natural language processing models based on the Transformer architecture can be directly applied to source code analysis tasks, such as code search. With the development of Transformer models designed more specifically for dealing with source code data, we believe the results on source code analysis tasks can be further improved.
قيم البحث
اقرأ أيضاً
To accelerate software development, developers frequently search and reuse existing code snippets from a large-scale codebase, e.g., GitHub. Over the years, researchers proposed many information retrieval (IR) based models for code search, which matc
h keywords in query with code text. But they fail to connect the semantic gap between query and code. To conquer this challenge, Gu et al. proposed a deep-learning-based model named DeepCS. It jointly embeds method code and natural language description into a shared vector space, where methods related to a natural language query are retrieved according to their vector similarities. However, DeepCS working process is complicated and time-consuming. To overcome this issue, we proposed a simplified model CodeMatcher that leverages the IR technique but maintains many features in DeepCS. Generally, CodeMatcher combines query keywords with the original order, performs a fuzzy search on name and body strings of methods, and returned the best-matched methods with the longer sequence of used keywords. We verified its effectiveness on a large-scale codebase with about 41k repositories. Experimental results showed the simplified model CodeMatcher outperforms DeepCS by 97% in terms of MRR (a widely used accuracy measure for code search), and it is over 66 times faster than DeepCS. Besides, comparing with the state-of-the-art IR-based model CodeHow, CodeMatcher also improves the MRR by 73%. We also observed that: fusing the advantages of IR-based and deep-learning-based models is promising because they compensate with each other by nature; improving the quality of method naming helps code search, since method name plays an important role in connecting query and code.
Code search is a common practice for developers during software implementation. The challenges of accurate code search mainly lie in the knowledge gap between source code and natural language (i.e., queries). Due to the limited code-query pairs and l
arge code-description pairs available, the prior studies based on deep learning techniques focus on learning the semantic matching relation between source code and corresponding description texts for the task, and hypothesize that the semantic gap between descriptions and user queries is marginal. In this work, we found that the code search models trained on code-description pairs may not perform well on user queries, which indicates the semantic distance between queries and code descriptions. To mitigate the semantic distance for more effective code search, we propose QueCos, a Query-enriched Code search model. QueCos learns to generate semantic enriched queries to capture the key semantics of given queries with reinforcement learning (RL). With RL, the code search performance is considered as a reward for producing accurate semantic enriched queries. The enriched queries are finally employed for code search. Experiments on the benchmark datasets show that QueCos can significantly outperform the state-of-the-art code search models.
To accelerate software development, much research has been performed to help people understand and reuse the huge amount of available code resources. Two important tasks have been widely studied: code retrieval, which aims to retrieve code snippets r
elevant to a given natural language query from a code base, and code annotation, where the goal is to annotate a code snippet with a natural language description. Despite their advancement in recent years, the two tasks are mostly explored separately. In this work, we investigate a novel perspective of Code annotation for Code retrieval (hence called `CoaCor), where a code annotation model is trained to generate a natural language annotation that can represent the semantic meaning of a given code snippet and can be leveraged by a code retrieval model to better distinguish relevant code snippets from others. To this end, we propose an effective framework based on reinforcement learning, which explicitly encourages the code annotation model to generate annotations that can be used for the retrieval task. Through extensive experiments, we show that code annotations generated by our framework are much more detailed and more useful for code retrieval, and they can further improve the performance of existing code retrieval models significantly.
In recent years, Neural Machine Translator (NMT) has shown promise in automatically editing source code. Typical NMT based code editor only considers the code that needs to be changed as input and suggests developers with a ranked list of patched cod
e to choose from - where the correct one may not always be at the top of the list. While NMT based code editing systems generate a broad spectrum of plausible patches, the correct one depends on the developers requirement and often on the context where the patch is applied. Thus, if developers provide some hints, using natural language, or providing patch context, NMT models can benefit from them. As a proof of concept, in this research, we leverage three modalities of information: edit location, edit code context, commit messages (as a proxy of developers hint in natural language) to automatically generate edits with NMT models. To that end, we build MODIT, a multi-modal NMT based code editing engine. With in-depth investigation and analysis, we show that developers hint as an input modality can narrow the search space for patches and outperform state-of-the-art models to generate correctly patched code in top-1 position.
There has been a growing interest in developing machine learning (ML) models for code learning tasks, e.g., comment generation and method naming. Despite substantial increase in the effectiveness of ML models, the evaluation methodologies, i.e., the
way people split datasets into training, validation, and testing sets, were not well designed. Specifically, no prior work on the aforementioned topics considered the timestamps of code and comments during evaluation (e.g., examples in the testing set might be from 2010 and examples from the training set might be from 2020). This may lead to evaluations that are inconsistent with the intended use cases of the ML models. In this paper, we formalize a novel time-segmented evaluation methodology, as well as the two methodologies commonly used in the literature: mixed-project and cross-project. We argue that time-segmented methodology is the most realistic. We also describe various use cases of ML models and provide a guideline for using methodologies to evaluate each use case. To assess the impact of methodologies, we collect a dataset of code-comment pairs with timestamps to train and evaluate several recent code learning ML models for the comment generation and method naming tasks. Our results show that different methodologies can lead to conflicting and inconsistent results. We invite the community to adopt the time-segmented evaluation methodology.
سجل دخول لتتمكن من نشر تعليقات
التعليقات
جاري جلب التعليقات
سجل دخول لتتمكن من متابعة معايير البحث التي قمت باختيارها
