自动生成EM的标题

IF 1.1 Q3 INFORMATION SCIENCE & LIBRARY SCIENCE

Digital Library Perspectives Pub Date : 2000-06-01 DOI:10.1145/336597.336670

Paul E. Kennedy, Alexander Hauptmann

{"title":"自动生成EM的标题","authors":"Paul E. Kennedy, Alexander Hauptmann","doi":"10.1145/336597.336670","DOIUrl":null,"url":null,"abstract":"Our prototype automatic title generation system inspired by statistical machine-translation approaches [1] treats the document title like a translation of the document. Titles can be generated without extracting words from the document. A large corpus of documents with human-assigned titles is required for training title \"translation\" models. On an f1 evaluation score our approach outperformed another approach based on Bayesian probability estimates [7].","PeriodicalId":42447,"journal":{"name":"Digital Library Perspectives","volume":"117 1","pages":"230-231"},"PeriodicalIF":1.1000,"publicationDate":"2000-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"11","resultStr":"{\"title\":\"Automatic title generation for EM\",\"authors\":\"Paul E. Kennedy, Alexander Hauptmann\",\"doi\":\"10.1145/336597.336670\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Our prototype automatic title generation system inspired by statistical machine-translation approaches [1] treats the document title like a translation of the document. Titles can be generated without extracting words from the document. A large corpus of documents with human-assigned titles is required for training title \\\"translation\\\" models. On an f1 evaluation score our approach outperformed another approach based on Bayesian probability estimates [7].\",\"PeriodicalId\":42447,\"journal\":{\"name\":\"Digital Library Perspectives\",\"volume\":\"117 1\",\"pages\":\"230-231\"},\"PeriodicalIF\":1.1000,\"publicationDate\":\"2000-06-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"11\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Digital Library Perspectives\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1145/336597.336670\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q3\",\"JCRName\":\"INFORMATION SCIENCE & LIBRARY SCIENCE\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Digital Library Perspectives","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1145/336597.336670","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"INFORMATION SCIENCE & LIBRARY SCIENCE","Score":null,"Total":0}

引用次数: 11

摘要

我们的原型自动标题生成系统受到统计机器翻译方法[1]的启发，将文档标题视为文档的翻译。无需从文档中提取单词就可以生成标题。训练标题“翻译”模型需要大量具有人工分配标题的文档语料库。在f1评估分数上，我们的方法优于另一种基于贝叶斯概率估计的方法[7]。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

查看原文本刊更多论文

Automatic title generation for EM

Our prototype automatic title generation system inspired by statistical machine-translation approaches [1] treats the document title like a translation of the document. Titles can be generated without extracting words from the document. A large corpus of documents with human-assigned titles is required for training title "translation" models. On an f1 evaluation score our approach outperformed another approach based on Bayesian probability estimates [7].

求助全文

通过发布文献求助，成功后即可免费获取论文全文。去求助

来源期刊

Digital Library Perspectives INFORMATION SCIENCE & LIBRARY SCIENCE-

CiteScore

3.90

自引率

11.80%

发文量

期刊介绍： Digital Library Perspectives (DLP) is a peer-reviewed journal concerned with digital content collections. It publishes research related to the curation and web-based delivery of digital objects collected for the advancement of scholarship, teaching and learning. And which advance the digital information environment as it relates to global knowledge, communication and world memory. The journal aims to keep readers informed about current trends, initiatives, and developments. Including those in digital libraries and digital repositories, along with their standards and technologies. The editor invites contributions on the following, as well as other related topics: Digitization, Data as information, Archives and manuscripts, Digital preservation and digital archiving, Digital cultural memory initiatives, Usability studies, K-12 and higher education uses of digital collections.