Relevance or Timeliness? The Real-Time Search Dilemma - Page 2
The challenge goes beyond filtering out spam and mere noise, although that is the first step of it. Social networking messages can lose their value within minutes and are generally considered non-authoritative. People who search real-time data, however, will increasingly raise their expectations and look for the same quality, authority, and relevance that they now look for when they perform traditional web searches.
Which algorithms will search engines use to ensure both relevance and timeliness?
Some may be straightforward, for example a twitter user who attracts many followers and whose tweets are often retweeted by other users might be prioritized as having more authority. A sudden spike in the prevalence of certain words - for example, flood or earthquake - may indicate an important event.
Connecting message content to geolocation could be another way to find relevance in specific circumstances related to updates about events or natural disasters.
But how will search engines deal with more subtle signs, like unusual phrasings or shifts in language in a commonly discussed topics, or other deviations which may signal fresh and interesting information versus a downpour of spam?
Further, real-time data needs to be indexed and updated instantaneously. Will search engines develop options allowing users to push new information, rather than waiting for the search engine crawling process? This could open web searches to new possibilities, like searching for real-time streams about the most updated traffic conditions at any moment, or the availability of parking spaces in a central area. How will relevance be assessed in these circumstances?
One thing is sure: we will be witnessing another interesting evolution in the future of search engines and internet search.