10 Natural Language Search Improvements Google Has Recently MadeBy: Chris Crum - August 13, 2012
Late on Friday, Google finally unveiled its “search quality highlights” lists for both June and July in one fell swoop. In all, there were 86 changes to wade through, and we’re still wading. You can see the whole list here.
1. One change (#82293), Google says, improves its dictionary search feature by adding support for more natural language searches. Unfortunately, it doesn’t get any more specific than that.
2. Another (timeob), Google says, improves natural language detection for the time feature, so that it will better understand questions like, ‘What time is it in India?’
3. Change #82496 involves the movie showtimes feature on mobile. Google says it improves recognition of natural language queries, as well as overall coverage.
4. One change (#82537) is the addition of natural language detection for currency conversion. This will enable Google to better understand queries like, ‘What is $500 in euros?,’ the company says.
5. A similar feature (#82519) has been added for the flight status feature so Google can better understand questions about flight arrival times and status.
6. With change #81776, Google says it has improved natural language detection for its unit conversion feature, sot that it will better understand questions like, “What is 5 miles in kilometers?”
7. Google says it has improved natural language processing for the dictionary search feature (#82887).
8. Google has made changes to the natural language detection for its sunrise/sunset feature (#82935).
9. Google says it has improved natural language detection for its MLB results, to better understand queries about baseball (schedules, scores, etc.). (#82536)
10. Finally, Google made another natural language-related change to the movie showtimes feature, claiming it has improved the natural language processing. (#82948)
All but number 10 were changes Google made in June. It appears that there was far more focus on natural language queries throughout that month than more recently.
Do you think Google has improved significantly in natural language search?