IRMA-International.org: Creator of Knowledge
Information Resources Management Association
Advancing the Concepts & Practices of Information Resources Management in Modern Organizations

Search Engine Result Bias: An Empirical Investigation of Commercial Web Based Search Tools

Search Engine Result Bias: An Empirical Investigation of Commercial Web Based Search Tools
View Free PDF
Author(s): Kholekile Gwebu (Kent State University, USA) and Jing Wang (Kent State University, USA)
Copyright: 2005
Pages: 4
Source title: Managing Modern Organizations Through Information Technology
Source Editor(s): Mehdi Khosrow-Pour (Information Resources Management Association, USA)
DOI: 10.4018/978-1-59140-822-2.ch142

Abstract

In recent years the world wide web has gained prominence as a prime resource for information on an array of topics including air travel, real estate and home décor just to name a few, which both individuals and organizations use to make informed decisions. Because the web is so vast and contains both structured and unstructured documents, web users often turn to web-based Information Retrieval Systems (IRS), typically referred to as search engines, as the main means of searching, sorting and navigating through the web. IRS in general have been in existence for decades and have allowed users to sort and search through structured documents, such as library records and news paper etc. IRS research tends to focus on the performance in terms of coverage, relevance, and ranking[1-6]. One major issue which has largely been ignored by researchers is that of bias in search engines. Bias simply refers to “undue inclusion or exclusion of certain items among those retrieved in response to queries or it is revealed in giving undue prominence to some items at the expense of others”[7]. Bias was previously never a serious issue in traditional IRS because the information being retrieved from them was not subject to systematic manipulation since it was largely non-commercial in nature. Today however, the competitive and commercial nature of search engines on the Web makes them vulnerable to systematic manipulation of results. Only a handful of studies devoted to assessing search engine bias are available on leading scholarly research databases [7], and even in such articles the authors have called for additional research into this area. If search results from leading search engines are indeed systematically skewed, web searchers need to be extremely cautious when attempting to retrieve fair and unbiased information from the web as some relevant search results obtained from search engines could intentionally be substituted with irrelevant but more commercially or politically appropriate results by search engine companies. This paper investigates the nature and extent of bias in commercial search engines. We consider the most popular search engines for assessment, as they are the ones which tend to have the most impact on internet users, then we use over 200 real user generated queries to assess bias across 8 different subject areas for all the search engines. The remainder of the paper is arranged as follows. The subsequent section synthesizes relevant literature on search engine bias. Thereafter, a set of hypotheses are presented followed by a description of the experiment conducted to assess bias, the empirical findings, and a discussion on those findings. Finally, we conclude by pointing out limitations of the study and issues which future researchers may wish to explore.

Body Bottom