Sonal Mahajan

Yusuke Kimura, Takumi Akazaki, Shinji Kikuchi et al.

Programmers often use Q&A sites (e.g., Stack Overflow) to understand a root cause of program bugs. Runtime exceptions is one of such important class of bugs that is actively discussed on Stack Overflow. However, it may be difficult for beginner programmers to come up with appropriate keywords for search. Moreover, they need to switch their attentions between IDE and browser, and it is time-consuming. To overcome these difficulties, we proposed a method, ``Q&A MAESTRO'', to find suitable Q&A posts automatically for Java runtime exception by utilizing structure information of codes described in programming Q&A website. In this paper, we describe a usage scenario of IDE-plugin, the architecture and user interface of the implementation, and results of user studies. A video is available at https://youtu.be/4X24jJrMUVw. A demo software is available at https://github.com/FujitsuLaboratories/Q-A-MAESTRO.

Ripon K. Saha, Akira Ura, Sonal Mahajan et al.

Automatic machine learning, or AutoML, holds the promise of truly democratizing the use of machine learning (ML), by substantially automating the work of data scientists. However, the huge combinatorial search space of candidate pipelines means that current AutoML techniques, generate sub-optimal pipelines, or none at all, especially on large, complex datasets. In this work we propose an AutoML technique SapientML, that can learn from a corpus of existing datasets and their human-written pipelines, and efficiently generate a high-quality pipeline for a predictive task on a new dataset. To combat the search space explosion of AutoML, SapientML employs a novel divide-and-conquer strategy realized as a three-stage program synthesis approach, that reasons on successively smaller search spaces. The first stage uses a machine-learned model to predict a set of plausible ML components to constitute a pipeline. In the second stage, this is then refined into a small pool of viable concrete pipelines using syntactic constraints derived from the corpus and the machine-learned model. Dynamically evaluating these few pipelines, in the third stage, provides the best solution. We instantiate SapientML as part of a fully automated tool-chain that creates a cleaned, labeled learning corpus by mining Kaggle, learns from it, and uses the learned models to then synthesize pipelines for new predictive tasks. We have created a training corpus of 1094 pipelines spanning 170 datasets, and evaluated SapientML on a set of 41 benchmark datasets, including 10 new, large, real-world datasets from Kaggle, and against 3 state-of-the-art AutoML tools and 2 baselines. Our evaluation shows that SapientML produces the best or comparable accuracy on 27 of the benchmarks while the second best tool fails to even produce a pipeline on 9 of the instances.

Sonal Mahajan, Mukul R. Prasad

Runtime Exceptions (REs) are an important class of bugs that occur frequently during code development. Traditional Automatic Program Repair (APR) tools are of limited use in this "in-development" use case, since they require a test-suite to be available as a patching oracle. Thus, developers typically tend to manually resolve their in-development REs, often by referring to technical forums, such as Stack Overflow (SO). To automate this manual process we extend our previous work, MAESTRO, to provide real-time assistance to developers for repairing Java REs by recommending a relevant patch-suggesting SO post and synthesizing a repair patch from this post to fix the RE in the developer's code. MAESTRO exploits a library of Runtime Exception Patterns (REPs) semi-automatically mined from SO posts, through a relatively inexpensive, one-time, incremental process. An REP is an abstracted sequence of statements that triggers a given RE. REPs are used to index SO posts, retrieve a post most relevant to the RE instance exhibited by a developer's code and then mediate the process of extracting a concrete repair from the SO post, abstracting out post-specific details, and concretizing the repair to the developer's buggy code. We evaluate MAESTRO on a published RE benchmark comprised of 78 instances. MAESTRO is able to generate a correct repair patch at the top position in 27% of the cases, within the top-3 in 40% of the cases and overall return a useful artifact in 81% of the cases. Further, the use of REPs proves instrumental to all aspects of MAESTRO's performance, from ranking and searching of SO posts to synthesizing patches from a given post. In particular, 45% of correct patches generated by MAESTRO could not be produced by a baseline technique not using REPs, even when provided with MAESTRO's SO-post ranking. MAESTRO is also fast, needing around 1 second, on average, to generate its output.