The International Conference on Mining Software Repositories (MSR) is the premier conference for data science (DS), machine learning (ML), and artificial intelligence (AI) in software engineering. There are vast amounts of data available in software-related repositories, such as source control systems, defect trackers, code review repositories, app stores, archived communications between project personnel, question-and-answer sites, CI build servers, package registries, and run-time telemetry. The MSR conference invites significant research contributions in which software data plays a central role. MSR Technical Track submissions using data from software repositories, either solely or combined with data from other sources, can take many forms, including: studies applying existing DS/ML/AI techniques to better understand the practice of software engineering, software users, and software behavior; empirically-validated applications of existing or novel DS/ML/AI-based techniques to improve software development and support the maintenance of software systems; and cross-cutting concerns around the engineering of DS/ML/AI-enabled software systems.

The 22nd International Conference on Mining Software Repositories will be held on April 28-29, 2025, in Ottawa, Canada.
Evaluation Criteria

We invite both full (maximum ten pages, plus two additional pages of references) as well as short (four pages, plus references) papers to the Research Track. Full papers are expected to describe new techniques and/or novel research results, to have a high degree of technical rigor, and to be evaluated scientifically. Short papers are expected to discuss controversial issues in the field or present interesting or thought-provoking ideas that are not yet fully developed. Submissions will be evaluated according to the following criteria:

    Soundness: This aspect pertains to how well the paper’s contributions — whether they involve new methodologies, applications of existing techniques to unfamiliar problems, empirical studies, or other research methods — address the research questions posed and are backed by a thorough application of relevant research procedures. For short papers, the expectation is for more limited evaluations given their narrower scope.
    Relevance: The extent to which the paper successfully argues or illustrates that its contributions help bridge a significant knowledge gap or tackle a crucial practical issue within the field of software engineering.
    Novelty: How original the paper’s contributions are in comparison to existing knowledge or how significantly they contribute to the current body of knowledge. Note that this doesn’t discourage well-motivated replication studies.
    Presentation: How well-structured and clear the paper’s argumentation is, how clearly the contributions are articulated, the legibility of figures and tables, and the adequacy of English language usage. All papers should comply with the formatting instructions provided.
    Replicability: The extent to which the paper’s claims can be independently verified through available replication packages and/or sufficient information included in the paper to understand how data was obtained, analyzed, and interpreted, or how a proposed technique works. All submissions are expected to adhere to the Open Science policy below.