So it seems to work in a basic form. Depending on the time I have to complete my chapter I will be working on:
- parsing the query results using python (SAX or DOM)
- building the UI (web interface)
- erm,test [I will at least get to this stage for the book submission]
- make the service agent based using AI so we have an autonomous collection of query results. I would like to try to capture keywords from a users daily google search session and run queries against the NASA data sets using these. This could be part of my Masters thesis at ITC maybe.
Update 29/09/2006
As originally presented by Jeremiah Grossman at Black Hat USA 2006[4],
it is possible to detect if someone has visited an arbitrary URL. Billy Hoffman[5], has today released proof-of-concept to show in practice how by using JavaScript we can detect the search queries a user has entered into arbitrary search on *any* search engine.
It parses the URL's stored in your browsers History cache and extracts the keywords. For example if you had sometime in the past looked on google for amazonas and birds your query would look like:
google.com/?searchhl=en&q=amazonas+birds
the code extracts 'amazonas and birds' from this query and/or any combination of them such as 'Amazonas','amazonas','Birds' or 'birds'. It is now possible to work out where a site visitor has been browsing and what they have been looking for. I have this code.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.
.