Just recently, Google launched an exciting new feature to help you find the best answers to your everyday questions from publisher content throughout the entire web. They call it sentence-compression algorithms.
An in-depth article explained how Google is now using a hand-fed AI that could interpret and create featured pieces in the Google search results. The best part is– it gives answers, not just your typical search results.
The famous search engine just went live on their desktop search results with this feature called to be “sentence-compression algorithms”. This distinctive attribute is able to pick up how to take a long paragraph or sentence from any relevant pages on the entire web and pull out the upshot, or mainly the information that you’re looking for, as what the article said.
To sum it up, Google is getting better and better at searching for contents on the web and taking out the particular set of information that directly answers the topic being asked.
The question is; how did they get this to work?
According to the article from Wired: To train the artificial Q&A brain of Google, Orr and company used old news stories, where machines begin to see how headlines serve as brief summaries of the lengthy articles that follow. For now, the giant company still needs its elite team of PhD linguists.
They not only exhibit sentence compression but literally label parts of speech in some ways that enables the neural nets to understand how the human language functions. With about 100 PhD linguists throughout the globe, it is expected for the team and this feature to continue to grow in the years to come.
With Google Assistant, Google Home, and the growth of featured snippets in the search results, it is no surprise that this famous search engine is continuously advancing their technology around this challenge.