Building on a previous study that succeeded in mapping business competition positions at an industry level using Web co-link
analysis, the current study attempted to improve Web co-link analysis by adding Web page content to obtain the mapping at
a particular market segment level. This method combines the ideas of Web content mining with Web structure mining. The method
was tested in the WiMAX sector of the telecommunication industry. Specifically, the keyword WiMAX was incorporated into queries
that searched for co-links to pairs of company Websites. Two sets of data were collected: one with the proposed method and
one with co-link search alone. The resulting two data matrices were analyzed using multidimensional scaling (MDS) to generate
maps of business competition. The comparison between the two maps shows that the proposed method produced a map focusing on
the WiMAX sector. The study also proposed the measure of reduction of co-link count that can be used to gauge the effectiveness of focusing the analysis on a particular sector. The reduction of co-link count could also be an easy and pragmatic measure for an analysis of a company’s competitiveness in a particular market segment.
Based on the findings from earlier studies which showed that links to business Websites contain useful business information,
we examined the feasibility of using Web co-link data to compare business competitive positions. We hypothesized that the
number of co-links to a pair of business Websites is a measure of the similarity between the two companies. Since similar
or related businesses are competing businesses, the co-link data can be used to map business competitive positions. We selected
32 telecommunications companies for the study and collected co-link data to these companies from Yahoo!. Multidimensional
scaling (MDS) analysis on the co-link data correctly mapped these companies into telecommunications industry sectors. This
proved our hypothesis and further confirmed the theory that links to business Websites can be objects for Web data mining.
We collected data in a way that would reflect two markets, the global market and the Chinese market. Results from the two
data sets revealed the competitive positions of the companies in the two markets. We propose that regular data collection
and analysis based on this method can be used to monitor the business competitive environment and trigger early warnings on
the change of the competitive landscape.