Categories
World Wide Web

A query language for searching websites

It does seem stale that you just write a set of words to query for a particular piece of information. For example, consider a query such as:

“Were there any video releases in the last one week that had XYZ(say an actor) as the lead role?”

Suppose we want to query this on normal search engines (what are abnormal search engines?!), we would probably write (disregarding useless advanced search options):

Search: “video release” XYZ -some-date-

Can we make it more specific? Probably we can, but the point is that there is no relationship between “video release” and XYZ whatsoever.

What could be a better query? How about using SQL-like queries to make our searches more specific?

It is possible and that is what semantic web, ontologies and a whole bag of technologies deal with. I am surely not going into the depths with these terminologies in this blog entry, but would like to hint on what modern searches could be like.

While modern English language interpretation would probably take a while to develop (we as SE's know the problem inherent in natural languages right?), it would surely be possible to develop a search engine which would be able to perform SQL-like queries.

For ex, it might be possible to put forth a query such as:

ALL(dom(video)) WHERE (video.leadrole=XYZ AND video.releasedate< -some-date- AND video.releaselocation=INDIA AND !video.actor=ABC) While whatever I have written above is directly off my mind (without any analysis whatsoever), it makes perfect sense right? It would mean: "Get me all the video releases which have XYZ as lead role and the release date is less than -some-date- and the release location is INDIA and it does not have ABC as an actor in it." The result? It might vary. While search engines might show a list of relevant sites as they do now, they could also take us directly to some particular relevant site as in "I am feeling lucky" of Google. (Then there would be more lucky people around 🙂 ) How's that? Amuzed? I surely am. And what's this going to use in the backend? XML of course. In fact there is an ontology language by name OWL, which is based on RDF, which is based on XML! And finally an update on the new search/RSS tools: A ticker from Yahoo – This ticker is able to present all the news feeds in box in your taskbar. So there is no need to view the RSS feeds in browsers. It is also able to show you the feeds present in your My-Yahoo page. Just let it tick…

I also hit upon another search engine called Mooter, which arranges the search results in clusters. Worth giving a try.

Eurekster – a group search engine, which learns from previously made queries.

And a last one in the list is, MyGoogleSearch, which is able to “guess” and extract keywords from a “similar” web-site. Not quite accurate, but worth giving a try.