Thursday, September 1, 2011

Same query Same results

I was trying to google something and was not able to find relevant results. Surprisingly I tried the same query again after a couple of days and got the same set of results. I was logged-in in gmail both the times. I was wondering why google search is not learning that I don't want these results; I want something else, something different.

When in last IR class (8/30/2011) we discussed the factors contributing to the relevance number R(.) of a document, I found that it very well depends upon the already retrieved documents. I questioned, then why google or any other search engine has not yet implemented it.

While I was googling about the same, I learned that it is difficult to completely remove the documents which are already shown, (and not judged by the user,) when the user re-enters the same query. This is because those documents are actually relevant and they are ranked based on their relevance to the query.
As we very well know that user is not sure what he is looking for, he is not inclined to read certain results. That doesn't imply that next same search should eliminate those documents. The maverick user might find the same set of results useful some other day (in some other light :-) ).

On the other hand, user might be aware of certain results and sure that they are definitely not relevant to his needs and hence never bothered to visit them. In such cases, he might find it frustrating to get the same results.

Hence, how to find "sweet spot" in such cases? If time permits, we can discuss this in tomorrow's class.

No comments:

Post a Comment