1 / 1

Motivation

An expert search query log The query log is extracted from Indiana Database of University Research Expertise (INDURE): https://www.indure.org The log is captured over a period of 90 days from October 19, 2010, to January 16, 2011 and it includes a series of requests

talib
Download Presentation

Motivation

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. An expert search query log • The query log is extracted from Indiana Database of University Research Expertise (INDURE): https://www.indure.org • The log is captured over a period of 90 days from October 19, 2010, to January 16, 2011 and it includes a series of requests • Each request includes the following fields: visitor IP, requested page, session ID, timestamp, and query terms • 16,434 unique visitors posted 14,503 queries with totally 104,030 requests; 851 empty queries and 2,923 unique queries of which 1,297 are advanced queries. The total number of sessions is 85,215. • Conclusions and future work • One of the first work on expert search query log analysis • The findings can help improve user interface design for expert search • There is an abundance of research opportunities in the query log analysis for expert search • We will exploit the query log to develop ranking models for expert search by learning from the implicit user feedback Our Goal –to understand the special characteristics of user behavior in expert search Analysis of An Expert Search Query LogYi Fang*, NaveenSomasundaram*, Luo Si*, JeongwooKo** and AdityaMathur* *Department of Computer Science, Purdue University, West Lafayette, IN**Google Inc, Mountain View, CA Motivation • Expert search has made rapid progress in modeling, algorithms and evaluations in the recent years • Very few work on analyzing how users interact with expert search systems • Expert search users have a specialized type of information need. It is desirable to understand their characteristics so that suitable IR systems can be provided for them Analysis Figure 1: Statistics of the INDURE query log. (a) Query length frequency; (b) Cumulative percentage of frequent queries; (c) Session length frequency; (d) Cumulative percentage of session duration . Table 1: Comparison of statistics among Web search, children search and the INDURE expert search. “N/A” means the statistics is not available in the reference Table 2: Top 8 most frequent expert search queries • Findings • Expert search users generally • issue shorter queries • Issue more common queries • use more advanced search features • have fewer queries in a session • than general Web search (as well as children search) users do

More Related