Exploring Web keyword analysis as an alternative to link analysis: a multi-industry case
- 484 Downloads
The study explored the feasibility of using Web keyword analysis as an alternative to link analysis and tested the feasibility in a multi-industry environment. The keyword is the organization’s name, in this case the company name. American companies from five industries were included in the study. The study found that the Web visibility of a company as measured by the number of Webpages on which the company name appears correlates with the company’s business measures (revenue, profits, and assets). The correlation coefficients are similar to that between the inlink counts and the business measures. This suggests that the keyword count (searched by the company name) could replace inlink count as an alternative indicator of some commonly used business measures. The co-word (the co-occurrence of the names of two companies on Webpages) count was used as a measure of the relatedness of the two companies. Multidimensional scaling (MDS) analysis was applied to the co-word matrices and generated MDS maps that showed relationships among companies in a multi-industry context. Keyword data were collected from three different types of Websites (general Websites, blog sites, and Web news sites) and results were compared. The study found blog sites to be the better source to collect data for this type of study. The comparison of MDS maps generated from co-link data and the blog co-word data showed that the co-word analysis is as effective as co-link analysis in mapping business relationships. The value of the study is not limited to the business sector as the co-word method could be applied to analysing relationships among other types of organizations.
KeywordsWeb keyword analysis Web visibility Business performance Business relationships Webometrics
The first author is supported by a research grant from the Social Sciences and Humanities Research Council of Canada (SSHRC) for the research program of Web data mining for business intelligence. Research assistant Qiang Ning helped with data collection. Part of the findings of the study were reported at the 7th International Conference on Webometrics, Informetrics and Scientometrics (WIS) & 12th COLLNET Meeting.
- Google (2011). Links to your site. http://www.google.com/support/webmasters/bin/answer.py?hl=en&answer=55281. Accessed 14 July 2011.
- Holmberg, K.J. (2009). Webometric network analysis: Mapping cooperation and geopolitical connections between local government administration on the Web. Ph.D. dissertation, Åbo Akademi University, Finland.Google Scholar
- Kim, J. H., Barnett, G. A., & Park, H. W. (2010). A Hyperlink and issue network analysis of the United States senate: A rediscovery of the Web as a relational and topical medium. Journal of the American Society for Information Science and Technology, 61(8), 1598–1611.Google Scholar
- Mergent (2009). North America—heavy construction sectors. March 2009, http://webreports.mergent.com. Accessed 14 September 2010.
- Mergent (2010a, b, c, d). North America—(Banking Sectors/IT & Technology Sectors/Media Sectors/Mining Sectors). April 2010. http://webreports.mergent.com. Accessed 14 September 2010.
- Romero-Frías, E., & Vaughan, L. (2010b). Patterns of Web linking to heterogeneous groups of companies: The case of stock exchange indexes. Aslib Proceedings New Information Perspectives, 62(2), 144–164.Google Scholar
- Seidman, E. (2007). We are flattered, but… Retrieved Oct. 24, 2011 from http://web.archive.org/web/20081219045957/http://blogs.msdn.com/livesearch/archive/2007/03/28/we-are-flattered-but.aspx. (Could not retrieve the original page. This is the archived page on Internet Archive).
- Vaughan, L., Yang, R., & Tang, J. (in press). Web co-word analysis for business intelligence in the Chinese environment. Aslib Proceedings: New Information Perspectives.Google Scholar
- Yahoo! (2011). Web Search APIs from Yahoo! Search. http://developer.yahoo.com/search/web/webSearch.html. Accessed 24 October 2011.