Bing Academic and Movie Search Intelligent Autocomplete

[engine]

Bing has said its autocomplete for academic and movie search is enhanced to improve the search experience for users that, perhaps, cannot remember the full details of the query. The autocomplete allows the researcher to construct highly structured queries.

The academic suggestions enables the users to investigate relationships between papers, authors, topics and publications via the large object graph.
Here are some examples of the type of academic query.

• Find all papers by an author
• Find a paper written by particular co-authors
• Find a paper about a specific topic presented at a conference
• Suggest titles or authors Bing Academic and Movie Search Intelligent Autocomplete [blogs.bing.com]

Similarly, with a movie search, the natural language queries through autocomplete helps users find, for example

• Movies by director
• Movies starring an actor in a particular genre
• Movies from a particular year starring a certain actor
• Movies starring a pair of actors

For both academic searches and movies, the understanding of the underlying domain is represented by a graph. The data is derived from the semantic graph that Bing uses to understand the world. This is stored in a format that allows us to look up information at runtime within milliseconds, thousands of times per second as the user types. This graph store allows us to look up exact matches, such as ‘tom cruise.’ It is also able to support the notion that both ‘tom c’ and ‘tom connor cruise’ refer to the same person.

Bing goes on the explain how the technology understands the intent, and generates suggestions based upon the user input. It won't give a response until it reaches a trigger point of understanding the intent.

An additional benefit is that, since Bing fully understands the query that is constructed, we know that results will be returned. This avoids potentially returning ‘dead-end queries’ with no results, grammatically incorrect or misspelled queries, which can be the case with more generalized language model based synthetic queries.

It's worth reading the whole piece as it shows how the system builds responses based upon an understanding beyond just a database of urls.

[ergophobe]

Thanks for the link.

understanding beyond just a database of urls.

For a while, I felt this was coming slower than I expected. Lately it finally seems to be accelerating. A while back I said something to you about "Links won't count in 10 years" and you laughed. I meant that to be aimed at the low end of what I think is possible, but as things accelerate, I'm not so sure. Links will always count as some sort of signal as long as there are links, but when I see things like this it shows me that they may become a minor ranking factor way sooner than ten years from now.