Synthetic question formulation
Briefly, embodiments disclosed herein may relate to formulating synthetic questions, such as in response to a search query, for example. Candidate synthetic questions may be presented to a user who may initiate a search at least in part by selecting one or more candidate synthetic questions, for example, in an embodiment.
1. A method, comprising:
generating, utilizing at least in part a processor of a computing device, a search results web page including one or more search results based at least in part on one or more user-provided query terms and further including one or more synthetic questions also based at least in part on the one or more user-provided query terms, wherein the generating the search results web page includes:
generating a set of candidate synthetic questions based at least in part on the one or more user-provided query terms,
ranking the set of candidate synthetic questions, performing a diversity function on at least a subset of the ranked set of candidate synthetic questions to generate a plurality of diversified candidate synthetic questions, and
estimating respective quality scores for the plurality of diversified candidate synthetic questions;
communicating the search results web page between the computing device and a user computing device; and
directing the user to one or more online community question answering sites at least in part via communication of one or more community question answering site web pages between the computing device and the user computing device at least in part in response to obtaining a user selection of a particular synthetic question of the one or more synthetic questions.
2. The method of claim 1 , wherein the directing the user to the one or more online community question answering sites further comprises:
submitting the particular synthetic question to the one or more online community question answering sites.
3. The method of claim 1 , wherein the ranking the set of candidate synthetic questions comprises ranking the set of candidate synthetic questions at least in part according to a machine learning-to-rank function or at least in part according to a natural language function, or a combination thereof.
4. The method of claim 3 , wherein the ranking the set of candidate synthetic questions at least in part according to the machine learning-to-rank function or at least in part according to the natural language function, or the combination thereof, comprises ranking the set of candidate synthetic questions based at least in part on part of speech tags or dependency trees, or a combination thereof, generated for individual synthetic questions of the set of candidate synthetic questions.
5. The method of claim 1 , wherein the generating the search results page including the one or more synthetic questions further comprises identifying one or more of the plurality of diversified candidate synthetic questions to have respective estimated quality scores that at least meet a specified threshold.
6. The method of claim 1 , wherein the diversity function comprises employing a first-word filter or an edit-distance filter, or a combination thereof, for the at least the subset of the ranked set of candidate synthetic questions.
7. The method of claim 5 , wherein the estimating the respective quality scores for the plurality of diversified candidate synthetic questions comprises offline training of a machine learning tool.
8. A method comprising:
generating, utilizing at least in part a processor of a computing device, a search results web page including one or more search results based at least in part on one or more user-provided query terms and further including one or more synthetic questions also based at least in part on the one or more user-provided query terms;
communicating the search results web page between the computing device and a user computing device;
directing the user to one or more online community question answering sites at least in part via communication of one or more community question answering site web pages between the computing device and the user computing device at least in part in response to obtaining a user selection of a particular synthetic question of the one or more synthetic questions; and
generating one or more additional synthetic questions for display to the user via the one or more online community question answering site web pages based at least in part on the obtaining the user selection of the particular synthetic question to encourage search refinement.
9. An article, comprising: a non-transitory storage medium having stored thereon instructions executable by a computing device to:
generate a search results web page for display to a user to include one or more search results based at least in part on one or more user-provided query terms and further to include one or more synthetic questions also based at least in part on the one or more user-provided query terms, wherein, to generate the search result web page, the storage medium having stored thereon further instructions executable by the computing device to:
generate a set of candidate synthetic questions based at least in part on the one or more user-provided query terms,
rank the set of candidate synthetic questions, and
identify one or more of the plurality of diversified candidate synthetic questions to have respective estimated quality scores that at least meet a specified threshold,
wherein, to generate the plurality of synthetic questions, the storage medium having stored thereon further instructions executable by the computing device to perform a diversity function on at least a subset of the ranked set candidate synthetic questions to generate a plurality of diversified candidate synthetic questions;
communicate between the computing device and a user computing device the search results web page; and
direct the user to one or more online community question answering sites at least in part via communication of one or more community question answering site web pages between the computing device and the user computing device at least in part in response to obtaining a user selection of a particular synthetic question of the one or more synthetic questions.
10. The article of claim 9 , wherein, to direct the user to the one or more online community question answering sites, the storage medium having stored thereon further instructions executable by the computing device to:
submit the particular synthetic question to the one or more online community question answering sites.
11. The article of claim 9 , wherein the storage medium having stored thereon further instructions executable by the computing device to rank the set of candidate synthetic questions at least in part in accordance with a machine learning-to-rank function or at least in part in accordance with a natural language function, or a combination thereof.
12. The article of claim 11 , wherein the storage medium having stored thereon further instructions executable by the computing device to rank the set of candidate synthetic questions based at least in part on part of speech tags or dependency trees, or a combination thereof, to be generated for individual questions of the set of candidate synthetic questions.
13. The article of claim 9 , wherein the storage medium having stored thereon further instructions executable by the computing device to generate one or more additional synthetic questions for display to the user via the one or more online community question answering site web pages based at least in part on the user selection of the particular synthetic question to encourage search refinement.
14. An apparatus, comprising:
means for generating at a computing device a search results web page for display to a user including one or more search results based at least in part on one or more user-provided query terms and further including one or more synthetic questions also based at least in part on the one or more user-provided query terms, wherein the means for generating the search results web page includes:
means for generating a set of candidate synthetic questions based at least in part on the one or more user-provided query terms,
means for ranking the set of candidate synthetic questions,
means for performing a diversity function on at least a subset of the ranked set of candidate synthetic questions to generate a plurality of diversified candidate synthetic questions, and
means for estimating respective quality scores for the plurality of diversified candidate synthetic questions;
means for communicating the search results web page between the computing device and a user computing device; and
means for directing the user to one or more online community question answering sites at least in part via communication of one or more community question answering site web pages between the computing device and the user computing device at least in part in response to obtaining a user selection of a particular synthetic question of the one or more synthetic questions.
15. The apparatus of claim 14 , wherein the means for directing the user to the one or more online community question answering sites further comprises:
means for submitting the particular synthetic question to the one or more online community question answering sites.