Monday, February 6, 2012

Dear Search Engines ... We Want MORE!


As mentioned in class last Monday, the four basic functionalities of search engines are Crawling, Indexing, Ranking and Display. Out of those four, the latter was pointed out to be “on the verge of a big change”. With this in mind, we decided that it would be interesting to discuss on the future of this side of search machines and more generally on how search machines are expected to change in the next years.

Beginning with the display and the interface design, the old-fashioned format of a list of pages, seems to be one of the first aspects that search engine companies try to alter. What is wanted instead of a linear list of independent guesses to the user's question, is something that looks more like a complete answer; something that could possibly combine all different kinds of rich media such as video and audio into the most suitable form for the specific query. Of course, the main challenge is that each kind of query has a different structure and a different set of attributes, resulting in a different most suitable display format.

A relevant area where researchers focus, is natural-language searching. What is wanted here is that you should be able to talk to a search engine in your voice without having to break everything down into keywords. It is true that alternative mining methods other than using the “incredibly limiting” ([2]) keyword search box, do exist but there is a need ([3]) for a search engine that combine all these existing search technologies.

But it is not only the Display that will change in future. As Google's VP of Engineering for the EMEA, N.Mattos points out ([1]): “one of the big trends is around getting the internet to be more local and getting it to be more personalized”. This connects to the Ranking functionality of search engines and introduces a “social aspect of ranking”, that influences the search results. Searching in the future will take advantage of the underlying user's social graph by imposing it into an algorithmic analysis, in order to make search more efficient and more relevant and to further refine a query or disambiguate. Of course, efforts in this direction are already being made; according to N. Mattos “Google+” is to Google a social network that “is going to create the social graph for us”.

The catalogue of “wants” that could be addressed to search machines is not limited by any means to the aforementioned ones. There are lots of other interesting aspects which we cannot include all here. We choose to conclude this short post with the words of N. Mattos on what are the three major areas of investments for Google at this time: “[...]we not only invest in the volume but also we invest in the diversity and how we can handle different types of documents. But we also invest in speed because, at the end of the day,[...]they still want to see those search results in a fraction of a second”.

[1]http://memeburn.com/2011/12/the-future-of-search-is-social-qa-with-googles-nelson-mattos/
[2]http://www.nature.com/nature/journal/v476/n7358/full/476025a.html
[3]http://nicholasscalice.com/2010/10/02/the-future-of-search-engines/
[4]http://googleblog.blogspot.com/2008/09/future-of-search.html

No comments:

Post a Comment