Question answering using entity references in unstructured data
AssigneeGoogle Inc.
InventorsDvir Keysar, Tomer Shmiel
Life of this patent
Estimating the term…
Abstract
Methods, systems, and computer-readable media are provided for collective reconciliation. In some implementations, a query is received, wherein the query is associated at least in part with a type of entity. Top-ranked search results are caused to be generated based at least in part on the query. An entity of the type of entity is identified in the top-ranked results and a response to the query is generated from the top-ranked results, the response including a reference to the entity that includes text distinct from the terms of the query.
Claims20
A method comprising:
- receiving a question query, wherein the question query is associated with a type of entity requested in the question query;
- generating a set of search results based on the question query, wherein each search result in the set contains unstructured text content;
- retrieving previously generated data associated with top-ranked search results in the set, wherein the previously generated data includes entity references found within the unstructured text content of the respective top-ranked search results, the entity references corresponding to entities of the type of entity;
- and
- generating and providing a natural language answer to the question query, wherein the natural language answer includes at least one term from the question query and text identifying an entity reference determined from the previously generated data of the top-ranked search results.
A system comprising:
- at least one processor;
- and
- memory storing instructions that, when executed by the at least one processor, cause the system to perform operations including:
- receiving a question query, wherein the question query is associated with a type of entity requested in the question query;
- generating a set of search results based on the question query, wherein each search result in the set contains unstructured text content;
- retrieving previously generated data associated with top-ranked search results in the set, wherein the previously generated data includes entity references found within the unstructured text content of the respective top-ranked search results, the entity references corresponding to entities of the type of entity;
- and
- generating and providing a natural language answer to the question query, wherein the natural language answer includes at least one term from the question query and text identifying an entity reference determined from the previously generated data of the top-ranked search results.
A non-transitory computer-readable medium storing instructions that, when executed by at least one processor of a computing system, cause the computing system to perform operations comprising:
- receiving a question query, wherein the question query is associated with a type of entity requested in the question query;
- generating a set of search results based on the question query, wherein each search result in the set contains unstructured text content;
- retrieving previously generated data associated with top-ranked search results in the set, wherein the previously generated data includes entity references found within the unstructured text content of the respective top-ranked search results, the entity references corresponding to entities of the type of entity;
- and
- generating and providing a natural language answer to the question query, wherein the natural language answer includes at least one term from the question query and text identifying an entity reference determined from the previously generated data of the top-ranked search results.
Dependencies are inferred from the claim text.