Paul Oh

University of Massachusetts Amherst, Amherst Center, Massachusetts, United States

Are you Paul Oh?

Claim your profile

Publications (1)0 Total impact

  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: INTRODUCTION In general, interactions with Web search engines could be characterized as "one size fits all". There is no representation of user preferences, search context, or the task context. Because of this obvious limitation of current search technology, personalization and context have been identified as major research challenges in a number of workshops. This work describes an ongoing effort toward automatically inferring knowledge about users and context from user queries. We report experiments that focus on the recognition of one context feature, namely reading level, and show that queries from users of different level groups can be effectively separated. Scientific readablility [4] has been studied for decades. Many readability indices have been developed to determine the reading levels of texts. Popular ones include Flesch-Kincaid [3], SMOG [5], and FOG [2] tests. A statistical approach has been proposed recently in [6]. A common characteristic among all these methods is th
    05/2004;