Sentiment detection as a ranking signal for reviewable entities

A non-transitory computer-readable storage medium encoded with computer program code, the computer program code comprising:

• computer program code for receiving a search query that references an entity type;
• computer program code for identifying search results that are associated with the entity type of the search query, each of the search results associated with a respective entity of a plurality of entities of the entity type;
• computer program code for ranking the search results based on entity ranking scores associated with the entities, wherein ranking a given search result of the search results is based on the entity ranking score of the entity associated with the given search result, and wherein the entity ranking score of the entity is based upon sentiment scores associated with review texts referencing the entity and values indicating likelihoods that the review texts reference the entity;
• and
• computer program code for providing the ranked search results responsive to the search query.

A system comprising:

• at least one processor for executing instructions in program modules, the instructions including instructions to:
• receive a search query that references an entity type;
• identify search results that are associated with the entity type of the search query, each of the search results associated with a respective entity of a plurality of entities of the entity type;
• rank the search results based on entity ranking scores associated with the entities, wherein the instructions to rank a given search result of the search results is based on the entity ranking score of the entity associated with the given search result, and wherein the entity ranking score of the entity is based upon sentiment scores associated with review texts referencing the entity and values indicating likelihoods that the review texts reference the entity;
• and
• provide the ranked entities as search results responsive to the search query.

A computer-implemented method, comprising:

• receiving a search query that references an entity type;
• identifying search results that are associated with the entity type of the search query, each of the search results associated with a respective entity of a plurality of entities of the entity type;
• ranking the search results based on entity ranking scores associated with the entities, wherein ranking a given search result of the search results is based on the entity ranking score of the entity associated with the search result, and wherein the entity ranking score of the entity is based upon sentiment scores associated with review texts referencing the entity and values indicating likelihoods that the review texts reference the entity;
• providing the ranked search results responsive to the search query;
• monitoring a plurality of user interactions with the given search result in response to additional search queries, wherein monitoring the plurality of user interactions with the given search result comprises monitoring at least one of click throughs for the given search result and times spent at a document associated with the given search result;
• and
• revising the entity ranking score of the entity based at least in part on the plurality of user interactions with the given search result.