Publication Type

Conference Proceeding Article

Publication Date

8-2012

Abstract

Consumers in e-commerce acquire information through search engines, yet to date there has been little empirical study on how users interact with the results produced by search engines. This is analogous to, but different from, the ever-expanding research on clickstreams, where users interact with static web pages. We propose a new network approach to analyzing search engine server log data. We call this searchstream data. We create graph representations based on the web pages that users traverse as they explore the search results that their use of search engines generates. We then analyze the graph-level properties of these search network graphs by conducting cluster analysis. We report preliminary evidence the presence of heterogeneity among users in terms of how they interact with search engines. This suggests that search engine users may not all benefit from the same functionality in the search engines they rely upon. We also offer additional evidence on the empirical regularities associated with a variety of relevant issues that arise in the business-to business (B2B) e-market context that we have studied.

Keywords

Big data, clickstreams, data mining, graph theory, keyword search, online markets, search behavior, searchstreams.

Discipline

Computer Sciences | Management Information Systems

Research Areas

Information Systems and Management

Publication

ICEC '12: Proceedings of the 14th Annual International Conference on Electronic Commerce

First Page

274

Last Page

275

ISBN

9781450311977

Identifier

10.1145/2346536.2346589

Publisher

ACM

City or Country

Singapore

Additional URL

http://dx.doi.org/10.1145/2346536.2346589

Share

COinS