Configuring and Assembling Information Retrieval Based Solutions for Software Engineering Tasks

Abstract

Information Retrieval (IR) approaches are used to leverage textual or unstructured data generated during the software development process to support various software engineering (SE) tasks (e.g., concept location, traceability link recovery, change impact analysis, etc.). Two of the most important steps for applying IR techniques to support SE tasks are preprocessing the corpus and configuring the IR technique, and these steps can significantly influence the outcome and the amount of effort developers have to spend for these maintenance tasks. We present the use of Genetic Algorithms (GAs) to automatically configure and assemble an IR process to support SE tasks. The approach named IR-GA determines the (near) optimal solution to be used for each step of the IR process without requiring any training. We applied IR-GA on three different SE tasks and the results of the study indicate that IR-GA outperforms approaches previously used in the literature, and that it does not significantly differ from an ideal upper bound that could be achieved by a supervised approach and a combinatorial approach

BoostNSift: A Query Boosting and Code Sifting Technique for Method Level Bug Localization

“…These techniques operate using a variety of approaches: textual or structural analysis of the source code [32], [35], [36], historic analysis of software change information [31], [34] and dynamic analysis of pass/fail testcase information [15]. As bug reports are usually formulated in natural language and the source code also includes large amounts of comments and identifiers, Information-Retrieval (IR) based bug localization techniques are frequently proposed for bug localization [3], [6], [8], [9], [13], [17], [26], [27], [29], [32]- [35], [37].…”

Design of a Machine Learning-based Approach for Fragment Retrieval on Models

“…According to the number of articles for each task, ML techniques are more commonly used for Bug Localization than for Traceability Link Recovery or for Feature Location. While 47% of the articles apply ML techniques for Bug [14], [50], [51], [52], [53], [54], [55], [56], [57], [58], [59] [60], [61], [62], [63], [64], [65], [66], [67], [68], [69], [70], [71], [72], [73], [74], [75], [76], [77], [78], [79], [80], [81], [82], [83], [84], [85] [10], [86], [87], [88], [89], [90], [91], [92], [93], [94], [95], [96], [97],…”

“…In this table, our articles are highlighted using bold font. Thanks to this table, we can see that genetic algorithms are not only used for our works [9,14], but they also are used for many other research works [10,54,55,87,89,90,91,94,99,62,67,79].…”

“…Several works rely on genetic algorithms guided by Latent Semantic Analysis [10,89,91,99]. Other works rely on genetic algorithms guided by Latent Dirichlet Allocation [54,89]. The rest of the works rely on genetic algorithms guided by fitness functions, which have been specifically defined for their works.…”

“…Genetic algorithm and programming [9], [14], [54], [55] [62], [67], [79] [10], [87], [89], [90], [91], [94], [99] Association rules [53], [56] [88], [92], [96] Others [9], [14], [57] [74], [77], [82], [84], [85] [97], [100], [101], [102] the Learning to Rank algorithms to guide a genetic algorithm and none of them target models as the main software artifact, as our work does.…”