Domain-specific negative media search techniques
In some implementations, systems and methods that are capable of customizing negative media searches using domain-specific search indexes are described. Data indicating a search query associated with a negative media search for an entity and a corpus of documents to be searched are obtained. Content from a particular collection of documents from among the corpus of documents is obtained and processed. Multiple scores for the entity are computed based on processing the content obtained from the collection of documents. The multiple scores are aggregated to compute a priority indicator that represents a likelihood that the collection of documents includes content that is descriptive of derogatory information.
1. A method performed by one or more computers, the method comprising:
receiving a search query for an entity, wherein the search query comprises at least (i) a text fragment representing a name of the entity and (ii) a domain-specific index that specifies a first collection of documents predetermined to be associated with a particular negative media domain;
identifying, using the domain-specific index, a set of documents from among the first collection of documents that are responsive to the search query;
determining a likelihood of the text fragment being present in a second collection of documents that are not associated with the particular negative media domain;
computing a commonality score for the entity based on the likelihood determined for the text fragment being present in the second collection of documents;
computing, for each document included within the set of documents, a reliability score representing a likelihood that information specified by a particular document is associated with the entity, wherein reliability scores for the set of documents are computed based on the commonality score;
selecting a subset of documents from the set of documents based on reliability scores computed for the set of documents, wherein each document included in the subset of documents has a reliability score satisfying a set of reliability criteria;
determining a priority indicator for the entity based on information specified by the subset of documents; and
enabling a user to perceive the priority indicator.
2. The method of claim 1 , wherein determining the priority indicator for the entity comprises:
computing a set of scores for the entity based on information specified by the subset of documents selected from the set of documents; and
computing the priority indicator for the entity based on combining the set of scores.
3. The method of claim 2 , wherein the computing the set of scores comprises computing a set of concept scores for the entity based on information specified by the subset of documents selected from the set of documents, wherein each concept score included in the set of concept scores represents a likelihood that the entity is associated with a predetermined negative attribute of a corresponding reference entity group of a plurality of reference entity groups predetermined to be associated with derogatory information.
4. The method of claim 3 , wherein the set of concept scores are computed based at least on determining that the entity is included in a list of sanctioned entities.
5. The method of claim 3 , wherein the computing the set of scores comprises computing a set of relevancy scores for the subset of documents selected from the set of documents, wherein each relevancy score included in the set of relevancy scores represents a likelihood that a document included in the subset of documents includes information descriptive of a set of predetermined negative attributes.
6. The method of claim 1 , wherein computing a reliability score for a document included in the set of documents comprises:
identifying one or more text fragments from the document;
determining one or more topics associated with each of the one or more text fragments; and
determining a likelihood that at least one of the one or more topics is associated with the entity.
7. The method of claim 1 , wherein:
the set of reliability criteria comprises a predetermined threshold associated with reliability scores computed for the set of documents; and
the subset of documents includes documents have reliability scores satisfying the predetermined threshold.
8. The method of claim 1 , wherein selecting the subset of documents from the set of documents comprises:
identifying multiple different subsets of documents from the set of documents;
computing, for each of the multiple different subsets of documents, a reliability score representing a likelihood that information specified by a particular subset of documents is associated with the entity; and
selecting the subset of documents based on reliability scores computed for the multiple different subsets of documents.
9. The method of claim 8 , further comprising:
obtaining a set of public records specifying information for an entity name associated with the entity;
determining a portion of the set of public records that uniquely correspond to the entity; and
wherein the reliability scores for the multiple subsets of documents are computed based on the portion of the set of public records that uniquely correspond to the entity.
10. The method of claim 1 , further comprising:
obtaining a set of public records specifying information for an entity name associated with the entity;
determining a portion of the set of public records that uniquely correspond to the entity; and
wherein the reliability scores for the set of documents are computed based on the portion of the set of public records that uniquely correspond to the entity.
11. A system comprising:
one or more computing devices; and
one or more non-transitory storage devices storing instructions that, when received by the one or more computing devices, cause the one or more computing devices to perform operations comprising:
receiving a search query for an entity, wherein the search query comprises at least (i) a text fragment representing a name of the entity and (ii) a domain-specific index that specifies a first collection of documents predetermined to be associated with a particular negative media domain;
identifying, using the domain-specific index, a set of documents from among the first collection of documents that are responsive to the search query;
determining a likelihood of the text fragment being present in a second collection of documents that are not associated with the particular negative media domain;
computing a commonality score for the entity based on the likelihood determined for the text fragment being present in the second collection of documents;
computing, for each document included within the set of documents, a reliability score representing a likelihood that information specified by a particular document is associated with the entity, wherein reliability scores for the set of documents are computed based on the commonality score;
selecting a subset of documents from the set of documents based on reliability scores computed for the set of documents, wherein each document included in the subset of documents has a reliability score satisfying a set of reliability criteria;
determining a priority indicator for the entity based on information specified by the subset of documents; and
enabling a user to perceive the priority indicator.
12. The system of claim 11 , wherein determining the priority indicator for the entity comprises:
computing a set of scores for the entity based on information specified by the subset of documents selected from the set of documents; and
computing the priority indicator for the entity based on combining the set of scores.
13. The system of claim 12 , wherein the computing the set of scores comprises computing a set of concept scores for the entity based on information specified by the subset of documents selected from the set of documents, wherein each concept score included in the set of concept scores represents a likelihood that the entity is associated with a predetermined negative attribute of a corresponding reference entity group of a plurality of reference entity groups predetermined to be associated with derogatory information.
14. The system of claim 13 , wherein the set of concept scores are computed based at least on determining that the entity is included in a list of sanctioned entities.
15. At least one non-transitory computer-readable storage device storing instructions that, when received by one or more computing devices, cause the one or more computing devices to perform operations comprising:
receiving a search query for an entity, wherein the search query comprises at least (i) a text fragment representing a name of the entity and (ii) a domain-specific index that specifies a first collection of documents predetermined to be associated with a particular negative media domain;
identifying, using the domain-specific index, a set of documents from among the first collection of documents that are responsive to the search query;
determining a likelihood of the text fragment being present in a second collection of documents that are not associated with the particular negative media domain;
computing a commonality score for the entity based on the likelihood determined for the text fragment being present in the second collection of documents;
computing, for each document included within the set of documents, a reliability score representing a likelihood that information specified by a particular document is associated with the entity, wherein reliability scores for the set of documents are computed based on the commonality score;
selecting a subset of documents from the set of documents based on reliability scores computed for the set of documents, wherein each document included in the subset of documents has a reliability score satisfying a set of reliability criteria;
determining a priority indicator for the entity based on information specified by the subset of documents; and
enabling a user to perceive the priority indicator.
16. The at least one non-transitory computer-readable storage of claim 15 , wherein determining the priority indicator for the entity comprises:
computing a set of scores for the entity based on information specified by the subset of documents selected from the set of documents; and
computing the priority indicator for the entity based on combining the set of scores.
17. The at least one non-transitory computer-readable storage of claim 16 , wherein the computing the set of scores comprises computing a set of concept scores for the entity based on information specified by the subset of documents selected from the set of documents, wherein each concept score included in the set of concept scores represents a likelihood that the entity is associated with a predetermined negative attribute of a corresponding reference entity group of a plurality of reference entity groups predetermined to be associated with derogatory information.
18. The at least one non-transitory computer-readable storage of claim 17 , wherein the set of concept scores are computed based at least on determining that the entity is included in a list of sanctioned entities.
19. The at least one non-transitory computer-readable storage of claim 17 , wherein the computing the set of scores comprises computing a set of relevancy scores for the subset of documents selected from the set of documents, wherein each relevancy score included in the set of relevancy scores represents a likelihood that a document included in the subset of documents includes information descriptive of a set of predetermined negative attributes.