Generating insightful connections between graph entities

A method comprising:

• receiving, by at least one processor, a query, the query being associated with a first entity from a data graph and a second entity from the data graph;
• generating, by the at least one processor, a search result for the query using an index for searching a document corpus;
• identifying at least one third entity in the data graph that shares a relationship with the first entity and the second entity, the relationship and the third entity being a matched connection;
• and
• including information from the matched connection in the search result.

A computer-implemented method comprising:

• receiving, using at least one processor, a query that corresponds to a first entity associated with a data graph;
• identifying, using the at least one processor, a second entity in the data graph connected to the first entity by a first path in the data graph and by a second path in the data graph, the first path having a single relationship between the first entity and the second entity, and the second path having two matching relationships that connect the first entity and the second entity via a third entity;
• and
• including information about the first path and the second path in a search result for the query.

A method comprising:

• receiving, using at least one processor, a query that corresponds to a first entity associated with a data graph;
• identifying, using the at least one processor, a second entity in the data graph connected to the first entity by a plurality of paths, each path of the plurality having two relationships that connect the first entity and the second entity via a third entity, wherein each path of the plurality of paths has the same two relationships but a different third entity;
• and
• including information about the first entity and the second entity in a search result for the query.

A system comprising:

• a data graph that links entities by relationships;
• and
• a query handling system that includes:
• at least one processor, and
• a memory storing instructions that, when executed by the at least one processor cause the query handling system to perform operations including:
• generating a first-degree entity graph for a particular entity in the data graph, the first-degree entity graph having first-degree paths each comprising a first-degree relationship and a first-degree entity,
• generating a second degree entity graph for the particular entity in the data graph, the second degree entity graph having second degree paths each comprising a first-second-degree relationship connecting to a first second-degree entity and a second second-degree relationship connecting the first second-degree entity to a second second-degree entity,
• for a particular second second-degree entity in the second-degree paths:
• assigning each second second-degree relationship to a group,
• determining an amount of second second-degree relationships in each group,
• determining that the amount meets a threshold, and
• storing data relating to the particular entity, the second second-degree entity, and the group in a memory for later retrieval in response to queries,
• for a particular first-degree entity in the first-degree paths:
• identifying at least one matching path by identifying a second-degree entity in the second-degree paths that matches the particular first-degree entity,
• determining whether the first second-degree relationship and the second second-degree relationship of the matching path are of the same type, and
• storing data relating to the particular entity and the matching paths for later retrieval in response to queries when the first second-degree relationship and the second second-degree relationship are of the same type.

A computer-implemented method comprising:

• retrieving, using at least one processor, a plurality of entities from a data graph that meet predetermined criteria for a collection;
• computing a fact-value for each of the plurality of entities, resulting in a set of fact-values;
• identifying an outlier value from the set of fact-values;
• storing an entity identifier, a collection identifier, and a fact identifier for the outlier value in a memory;
• receiving a query that relates to a particular entity;
• determining whether an identifier for the particular entity matches the stored entity identifier for the outlier value;
• and
• including information relating to the outlier value in a result list for the query when it is determined that the identifier for the particular entity matches the stored entity identifier for the outlier value.