课题基金 / 基金详情

CAREER: Contextual Link Analysis

CAREER: Contextual Link Analysis
职业:情境链接分析
批准号:
0545875
负责人:
Brian Davison
金额:
$50.0万
依托单位:
依托单位国家:
美国
项目类别:
Continuing Grant
财政年份:
2006
资助国家:
美国
项目状态:
已结题
起止时间:
2006-08-15 至 2012-07-31

项目摘要

项目成果

Brian Davison的其他基金

相似基金

相关文献

中文摘要
翻译
本项目提案的重点是主题链接分析的问题,目的是提高网络搜索结果的相关性。一个网页(或作者)并不是在所有的主题上都具有同等的权威性。 然而,传统的链接分析技术只考虑单一的权威度量,很少或根本不关心权威来源的主题或社区。 这项研究找到了确定公认权威的背景,并将该背景作为时事性或社区认可纳入更准确的声誉自动分析的方法。 该项目通过区分流行度、质量和权威性进一步改进了链接分析。 通过将这种分析集成到基本的链接分析技术中,使用人类相关性和质量判断来证明搜索结果质量的改进。 这种上下文链接分析提供了新的机会,以显着改善测量的权威性,从而搜索引擎结果排名(以及为其他目的的主题排名),并更好地了解互联网和学术出版物等互联文件网络的主题结构。 该项目还包括教育和推广活动,包括一个计划,以激发本科生的兴趣和经验,在计算机科学研究领域的学校,同时促进附近的教师之间的互动。该项目网页(http://www.cse.lehigh.edu/tubrian/nsf/career-05.html)提供了研究结果和关于该项目的其他信息。
英文摘要
This project proposal focuses on the problem of topical link analysis with a goal to improve the relevance of web search results. A web page (or author) is not equally authoritative on all topics. However, traditional link analysis techniques consider only a single measure of authority, with little or no concern for the topic or community from which that authority is derived. This research finds ways to determine the context of recognized authority, and to incorporate that context as topicality or community recognition into more accurate automated analyses of reputation. This project further improves link analysis by distinguishing between popularity, quality, and authority. By integrating such analysis into fundamental link analysis techniques, search result quality improvements are demonstrated using human relevance and quality judgments. This contextual link analysis provides new opportunities to significantly improve measurements of authority and thus search engine result rankings (as well as topical rankings for other purposes), and to better understand the topical structure of interlinked document networks such as the Web and scholarly publications. The project additionally incorporates educational and outreach activities, including a program to stimulate undergraduate interest and experience in computer science research in area schools, while promoting interaction among nearby faculty. The project web page (http://www.cse.lehigh.edu/~brian/nsf/career-05.html) provides access to research results and additional information on project.
期刊论文(0)
专著(0)
科研奖励(0)
会议论文
III: Small: Collaborative Research: Algorithms, systems, and theories for exploiting data dependencies in crowdsourcing
  • 批准号:
    2008155
  • 项目类别:
    Standard Grant
  • 资助金额:
    $25.0万
  • 财政年份:
    2020
  • 负责人:
    Brian Davison
  • 依托单位:
REU Site: Intelligent and Scalable Systems
  • 批准号:
    1757787
  • 项目类别:
    Standard Grant
  • 资助金额:
    $36.0万
  • 财政年份:
    2018
  • 负责人:
    Brian Davison
  • 依托单位:
III: Small: Domain-Agnostic Dataset Search
  • 批准号:
    1816325
  • 项目类别:
    Standard Grant
  • 资助金额:
    $51.58万
  • 财政年份:
    2018
  • 负责人:
    Brian Davison
  • 依托单位:
III-COR-Medium: Efficient and Effective Search Services Over Archival Webs
  • 批准号:
    0803605
  • 项目类别:
    Standard Grant
  • 资助金额:
    $90.0万
  • 财政年份:
    2008
  • 负责人:
    Brian Davison
  • 依托单位:
海外基金