{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/learning-type-annotation-is-big-data-enough","title":"Learning Type Annotation: Is Big Data Enough?","arxiv_id":null,"date":"2021-08-18","proceeding":"ACM Joint Meeting on European Software Engineering Conference and Symposium on the Foundations of Software Engineering 2021 8","authors":["Kevin Jesse","Premkumar T. Devanbu","Toufique Ahmed"],"abstract":"TypeScript is a widely used optionally-typed language where developers can adopt “pay as you go” typing: they can add types as\r\ndesired, and benefit from static typing. The “type annotation tax”\r\nor manual effort required to annotate new or existing TypeScript\r\ncan be reduced by a variety of automatic methods. Probabilistic\r\nmachine-learning (ML) approaches work quite well. ML approaches\r\nuse different inductive biases, ranging from simple token sequences\r\nto complex graphical neural network (GNN) models capturing syntax and semantic relations. More sophisticated inductive biases are\r\nhand-engineered to exploit the formal nature of software. Rather\r\nthan deploying fancy inductive biases for code, can we just use “big\r\ndata” to learn natural patterns relevant to typing? We find evidence\r\nsuggesting that this is the case. We present TypeBert, demonstrating that even with simple token-sequence inductive bias used in\r\nBERT-style models and enough data, type-annotation performance\r\nof the most sophisticated models can be surpassed.","url_abs":"https://dl.acm.org/doi/abs/10.1145/3468264.3473135","url_pdf":"https://www.cs.ucdavis.edu/~devanbu/typebert_esec_fse_.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"learning-type-annotation-is-big-data-enough","repo_url":"https://github.com/typebert/typebert","is_official":0,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"tf","reach":null}],"tasks":[{"task_slug":"inductive-bias","task_name":"Inductive Bias"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}