{"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/building-large-machine-reading-comprehension","title":"Building Large Machine Reading-Comprehension Datasets using Paragraph Vectors","arxiv_id":"1612.04342","date":"2016-12-13","proceeding":null,"authors":["Radu Soricut","Nan Ding"],"abstract":"We present a dual contribution to the task of machine reading-comprehension:\na technique for creating large-sized machine-comprehension (MC) datasets using\nparagraph-vector models; and a novel, hybrid neural-network architecture that\ncombines the representation power of recurrent neural networks with the\ndiscriminative power of fully-connected multi-layered networks. We use the\nMC-dataset generation technique to build a dataset of around 2 million\nexamples, for which we empirically determine the high-ceiling of human\nperformance (around 91% accuracy), as well as the performance of a variety of\ncomputer models. Among all the models we have experimented with, our hybrid\nneural-network architecture achieves the highest performance (83.2% accuracy).\nThe remaining gap to the human-performance ceiling provides enough room for\nfuture model improvements.","url_abs":"http://arxiv.org/abs/1612.04342v1","url_pdf":"http://arxiv.org/pdf/1612.04342v1.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":"building-large-machine-reading-comprehension","repo_url":"https://github.com/google/mcafp","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":null}],"tasks":[{"task_slug":"dataset-generation","task_name":"Dataset Generation"},{"task_slug":"machine-reading-comprehension","task_name":"Machine Reading Comprehension"},{"task_slug":"reading-comprehension","task_name":"Reading Comprehension"}],"methods":[],"datasets_introduced":[{"slug":"mc-afp","name":"MC-AFP","full_name":""}],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}