{"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/asymmetric-minwise-hashing","title":"Asymmetric Minwise Hashing","arxiv_id":"1411.3787","date":"2014-11-14","proceeding":null,"authors":["Anshumali Shrivastava","Ping Li"],"abstract":"Minwise hashing (Minhash) is a widely popular indexing scheme in practice.\nMinhash is designed for estimating set resemblance and is known to be\nsuboptimal in many applications where the desired measure is set overlap (i.e.,\ninner product between binary vectors) or set containment. Minhash has inherent\nbias towards smaller sets, which adversely affects its performance in\napplications where such a penalization is not desirable. In this paper, we\npropose asymmetric minwise hashing (MH-ALSH), to provide a solution to this\nproblem. The new scheme utilizes asymmetric transformations to cancel the bias\nof traditional minhash towards smaller sets, making the final \"collision\nprobability\" monotonic in the inner product. Our theoretical comparisons show\nthat for the task of retrieving with binary inner products asymmetric minhash\nis provably better than traditional minhash and other recently proposed hashing\nalgorithms for general inner products. Thus, we obtain an algorithmic\nimprovement over existing approaches in the literature. Experimental\nevaluations on four publicly available high-dimensional datasets validate our\nclaims and the proposed scheme outperforms, often significantly, other hashing\nalgorithms on the task of near neighbor retrieval with set containment. Our\nproposal is simple and easy to implement in practice.","url_abs":"http://arxiv.org/abs/1411.3787v1","url_pdf":"http://arxiv.org/pdf/1411.3787v1.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":"asymmetric-minwise-hashing","repo_url":"https://github.com/ritchie46/lsh-rs","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"retrieval","task_name":"Retrieval"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}