| //--> |
By Nicholas Carr
How do you parse the tweet? Five years ago, that question would be nonsense. Today, perfectly reasonable, and that's in front of Amit Singhal's minds. Singhal Google search led to insert new data into search results in real time by tracking and ranking updates online content - especially the thousands of messages which of course through the social network every second.
real-time search is a response to fundamental changes in how people use the Web. People used to visit the page, click on the link, and visit other pages. Now they spend much time monitoring of data streams - tweets, status update, headlines - from services like Facebook and Twitter, and also from blogs and news outlets.
Brief info-nuggets are the new currency the Web, and sort through them for useful information is a challenge for search engines. Its the most frightening aspect, according to Singhal, does not collect data. Facebook and Twitter are happy to sell access to their data feed - or "fire hose," which they call - directly to the search service provider, the information that flows directly into Google's computer.
What's really hard about the real-time search is to find out the meaning and value of quick bits of information. The challenge goes beyond spam filtering, though it is an important part of it. People who are looking for real-time data to the same quality, authority, and relevance of what they expect when they engage in traditional Web search. Nobody wants to drink directly from a fire hose.
Google dominates traditional search carefully tracking signal and links to other pages in value as they accumulate over time. But the search for real-time, this does not work. Social-networking message can lose their value within minutes after the written. Google has to measure their value in seconds, or even microseconds.
Source:
http://www.technologyreview.com/computing/25079/
No comments:
Post a Comment