Web & Text mining
For BNOSAC, Text & Web mining has shown its use in the following settings:
Extract information from competitors from the web
We recently developed Python scripts to extract the prices of competitor products from the web.
We can also set up scripts that fetch other information from the web overnight.
Sentiment analysis
We have already successfully set up web farms that collected huge amount of information from web blogs. For that purpose, sites like Technorati were a great help. The collected text was subsequently fed to text mining algorithms to analyse it's content for relevant information. You can contact us for more information.
Web log analysis
We are currently analysing publicly available web logs to see if we can somehow incorporate recommender systems on a website so that customers who visit a site can more efficiently be redirected to what they are really interested in.
