The effect of title term suggestion on e-commerce sites

Most E-Commerce websites rely on title keyword search to accurately retrieve the items for sale in a particular category. We have found that the titles of many items on eBay are shortened or not very specific, which leads to ineffective results when searched. One possible solution is to recommend the sellers relevant and informative terms for title expansion without any change of search function. The related technique has been explored in previous work such as query expansion and keyword suggestion. In this paper, we study the effect of term suggestion on title-based search. A frequently used approach, co-occurrence, is tested on a dataset collected from eBay website (www.ebay.com). Besides, for suggestion algorithm, we take into account three particular features in our application scenario, including concept term, description relevance and chance-to-be viewed. Although the experiments are conducted on eBay data, we believe that considering E-Commerce particularities will help us to customize the suggestion according to the requirements of web commerce.

[1]  Charles L. A. Clarke,et al.  Shortest Substring Ranking (MultiText Experiments for TREC-4) , 1995, TREC.

[2]  Gerard Salton,et al.  Term-Weighting Approaches in Automatic Text Retrieval , 1988, Inf. Process. Manag..

[3]  James Allan,et al.  Automatic Query Expansion Using SMART: TREC 3 , 1994, TREC.

[4]  Jacques Savoy,et al.  Term Proximity Scoring for Keyword-Based Retrieval Systems , 2003, ECIR.

[5]  Gerard Salton,et al.  Automatic Text Processing: The Transformation, Analysis, and Retrieval of Information by Computer , 1989 .

[6]  Wei-Ying Ma,et al.  Probabilistic query expansion using query logs , 2002, WWW '02.

[7]  Joshua Goodman,et al.  Finding advertising keywords on web pages , 2006, WWW '06.

[8]  Francine Chen,et al.  A trainable document summarizer , 1995, SIGIR '95.

[9]  Wei-Ying Ma,et al.  Learning to cluster web search results , 2004, SIGIR '04.

[10]  W. Bruce Croft,et al.  Improving the effectiveness of information retrieval with local context analysis , 2000, TOIS.

[11]  Yiming Yang,et al.  An Evaluation of Statistical Approaches to Text Categorization , 1999, Information Retrieval.

[12]  Peter Willett,et al.  The Limitations of Term Co-Occurrence Data for Query Expansion in Document Retrieval Systems , 1991 .

[13]  Gerard Salton,et al.  Research and Development in Information Retrieval , 1982, Lecture Notes in Computer Science.

[14]  Xiaoyuan Wu,et al.  Keyword extraction for contextual advertisement , 2008, WWW.

[15]  Kenneth Ward Church,et al.  Word Association Norms, Mutual Information, and Lexicography , 1989, ACL.

[16]  Shuming Shi,et al.  Web page title extraction and its application , 2007, Inf. Process. Manag..

[17]  Alexander G. Hauptmann,et al.  Automatic title generation for EM , 2000, DL '00.

[18]  Alexander Hauptmann,et al.  Automatic Title Generation using EM , 2000 .

[19]  Rong Jin,et al.  A New Probabilistic Model for Title Generation , 2002, COLING.

[20]  Chris Buckley,et al.  Improving automatic query expansion , 1998, SIGIR '98.

[21]  W. Bruce Croft,et al.  Discovering key concepts in verbose queries , 2008, SIGIR '08.

[22]  Berthier A. Ribeiro-Neto,et al.  Impedance coupling in content-targeted advertising , 2005, SIGIR '05.

[23]  Yifan Chen,et al.  Advertising keyword suggestion based on concept hierarchy , 2008, WSDM '08.

[24]  Peter Willett,et al.  The limitations of term co-occurrence data for query expansion in document retrieval systems , 1991, J. Am. Soc. Inf. Sci..