An incremental update framework for efficient retrieval from software libraries for bug localization

Abstract

Abstract-Information Retrieval (IR) based bug localization techniques use a bug reports to query a software repository to retrieve relevant source files. These techniques index the source files in the software repository and train a model which is then queried for retrieval purposes. Much of the current research is focused on improving the retrieval effectiveness of these methods. However, little consideration has been given to the efficiency of such approaches for software repositories that are constantly evolving. As the software repository evolves, the index creation and model learning have to be repeated to ensure accuracy of retrieval for each new bug. In doing so, the query latency may be unreasonably high, and also, re-computing the index and the model for files that did not change is computationally redundant. We propose an incremental update framework to continuously update the index and the model using the changes made at each commit. We demonstrate that the same retrieval accuracy can be achieved but with a fraction of the time needed by current approaches. Our results are based on two basic IR modeling techniques -Vector Space Model (VSM) and Smoothed Unigram Model (SUM). The dataset we used in our validation experiments was created by tracking commit history of AspectJ and JodaTime software libraries over a span of 10 years.

Comparing Incremental Latent Semantic Analysis Algorithms for Efficient Retrieval from Software Libraries for Bug Localization

“…Change Processing: For each new commit, the set of source files that are affected, called the change-set, is checked out and subject to the text preprocessing steps described in [10].…”

“…Similar to our previous work [10], we have used the moreBugs [23] dataset to perform our experimental validation. The dataset contains all the necessary information to evaluate both the batch-mode and the incremental-mode approaches to IR based bug localization, namely: (a) the commit-level changes taking place in the repository; (b) the release history of the software; and (c) a set of closed/resolved issues/bugs.…”

“…Subsequently, a text model is learned from this index. A bug is localized by treating the textual content of the bug as a query that is then used to fetch the relevant documents from the model [10].…”

“…• Since it is often the case that each commit to a repository affects only a very tiny portion of the overall code base [10], re-learning the model from scratch for each new bug is inefficient and may be unnecessarily computationally expensive.…”

“…In a previous contribution [10], we presented an incremental update framework that keeps the index and the model updated at each commit as the software evolves using the Smoothed Unigram Model (SUM) and the Vector Space Model (VSM). That work demonstrated that the SUM and VSM based incremental frameworks gave us enhanced retrieval efficiency with no penalty in retrieval accuracy.…”