Methods and apparatus for searching for content and providing content of interest to one or more users
Methods and system for providing information selected from a large set of digital content to at least one user. One such method comprises receiving user context information associated with the at least one user and identifying or generating, using at least one processor executing stored program instructions, a first concept in a semantic network, the first concept representing at least a portion of the user context information. The method further comprises obtaining at least one concept, including a second concept, semantically relevant to the first concept at least in part, by synthesizing the second concept based on the first concept and at least one other concept in the semantic network; and providing information to the at least one user, wherein the information is selected by using the first concept and the at least one obtained concept semantically relevant to the first concept, wherein the first concept in a semantic network is represented by a data structure storing any data associated with a node in the semantic network.
1 . A computer-implemented method for improving querying techniques used in providing to a search service textual information that is personalized to at least one user, the method comprising executing via a computer system comprising at least one processor and at least one non-transitory memory:
receiving, by the processor, user context information including textual information indicative of the interests associated with the at least one user;
identifying or generating, by the processor using a semantic analysis module, an active concept in a semantic network encoded in non-transitory memory, the active concept representing at least a portion of the user context information;
augmenting, by the processor, the user context information using a semantic analysis operations comprising:
retrieving, by the processor, one or more concepts from a target knowledge representation and one or more concepts from a reference knowledge representation, the target knowledge representation and the reference knowledge representation encoded in non-transitory computer-readable memory, wherein the target knowledge representation represents knowledge from a target set of digital content indexed by a search engine, and wherein the reference knowledge representation represents a reference set of digital content;
analyzing, by the processor, the one or more concepts from the target knowledge representation and the one or more concepts from the reference knowledge representation to identify sets of target-attributes and reference-attributes, wherein the target-attributes and reference-attributes are more fundamental knowledge representation entities that comprise the one or more concepts from the target knowledge representation and the one or more concepts from the reference knowledge representation, respectively;
synthesizing, by the processor, a merged knowledge representation derived from merging the target knowledge representation and the reference knowledge representation based on the sets of target-attributes and reference-attributes around one or more concepts in common between the target knowledge representation and the reference knowledge representation, the merged knowledge representation encoded in non-transitory memory and comprising a plurality of concepts and attributes semantically relevant to the active concept;
determining, by the processor, a semantic relevance score for the plurality of concepts and attributes semantically relevant to the active concept using a measure of relevance, the semantic relevance score for each concept or each attribute of the plurality of concepts and attributes being indicative of semantic relevance of that concept or attribute to the active concept, and wherein the semantic relevance score is determined based at least in part on the structure of the merged knowledge representation; and
constructing, by the processor, one or more search queries derived from the user context information and labels of the plurality of concepts and attributes semantically relevant to the active concept, based on the semantic relevance score computed for each concept or attribute of the plurality of concepts and attributes semantically relevant to the active concept; and
returning the one or more search queries to the search service to retrieve textual information personalized to the at least one user.
2 . The computer-implemented method of claim 1 , wherein synthesizing the merged knowledge representation comprises pruning one or more concepts from the reference knowledge representation and/or from the target knowledge representation prior to the merging.
3 . The computer-implemented method of claim 1 , wherein synthesizing the merged knowledge representation comprises using an addition operation based on an analogy-by-parent technique and/or an analogy-by-sibling technique.
4 . The computer-implemented method of claim 1 , wherein synthesizing the merged knowledge representation comprises using a substitution operation, wherein the substitution operation comprises using a retrieval operation or an addition operation.
5 . The computer-implemented method of claim 1 , wherein computing the semantic relevance score for a concept or attribute in the plurality of concepts and attributes semantically relevant to the active concept comprises using at least one measure of relevance from among generation certainty, concept productivity, Jaccard, statistical coherence, and cosine similarity.
6 . The computer-implemented method of claim 1 , wherein identifying or generating the active concept comprises:
determining whether the user context information matches an identifier of a concept in the merged knowledge representation; and
when it is determined that the user context information does not match the identifier of the concept in the merged knowledge representation, generating the active concept in the merged knowledge representation.
7 . The computer-implemented method of claim 1 , wherein the user context information comprises at least one of demographic information associated with the user, information from the user's browsing history, information typed in by the user, and information highlighted by the user.
8 . The computer-implemented method of claim 1 , wherein the target knowledge representation is represented by a data structure embodying a directed graph comprising a plurality of nodes and a plurality of edges, wherein each node is associated with a concept and an edge between two nodes represents a relationship between the two corresponding concepts.
9 . The computer-implemented method of claim 1 , wherein the target set of content comprises content accessible through an e-commerce website, a website of a business, a website providing access to one or more databases, an online portal, or a corporate Intranet.
10 . The computer-implemented method of claim 1 , wherein the reference set of digital content comprises content accessible through an information repository, database, content-provisioning source, or reference corpora.
11 . The computer-implemented method of claim 1 , wherein the user context information comprises a user-specified request or query.
12 . A system for improving querying techniques used in providing to a search service textual information that is personalized to at least one user, the system comprising:
at least one processor, at one or more server computers, configured to perform a method comprising:
receiving user context information including textual information indicative of the interests associated with the at least one user;
identifying or generating, using a semantic analysis module, an active concept in a semantic network encoded in non-transitory memory, the active concept representing at least a portion of the user context information;
augmenting the user context information using a semantic analysis operations comprising:
retrieving one or more concepts from a target knowledge representation, the target knowledge representation and the reference knowledge representation encoded in non-transitory computer-readable memory, wherein the target knowledge representation represents knowledge from a target set of digital content indexed by a search engine, and wherein the reference knowledge representation represents a reference set of digital content;
analyzing the one or more concepts from the target knowledge representation and the one or more concepts from the reference knowledge representation to identify sets of target-attributes and reference-attributes, wherein the target-attributes and reference-attributes are more fundamental knowledge representation entities that comprise the one or more concepts from the target knowledge representation and the one or more concepts from the reference knowledge representation, respectively;
synthesizing a merged knowledge representation derived from merging the target knowledge representation and the reference knowledge representation based on the sets of target-attributes and reference-attributes around one or more concepts in common between the target knowledge representation and the reference knowledge representation, the knowledge representation encoded in non-transitory memory and comprising a plurality of concepts and attributes semantically relevant to the active concept;
determining a semantic relevance score for the plurality of concepts and attributes semantically relevant to the active concept using a measure of relevance, the semantic relevance score for each concept or each attribute of the plurality of concepts and attributes being indicative of semantic relevance of that concept or attribute to the active concept, and wherein the semantic relevance score is determined based at least in part on the structure of the merged knowledge representation; and
constructing one or more search queries derived from the user context information and labels of the plurality of concepts and attributes semantically relevant to the active concept, based on the semantic relevance score computed for each concept or attribute of the plurality of concepts and attributes semantically relevant to the active concept; and
returning the one or more search queries to the search service to retrieve textual information personalized to the at least one user.
13 . The system of claim 12 , wherein synthesizing the merged knowledge representation comprises pruning one or more concepts from the reference knowledge representation and/or from the target knowledge representation prior to the merging.
14 . The system of claim 12 , wherein synthesizing the merged knowledge representation comprises using an addition operation based on an analogy-by-parent technique and/or an analogy-by-sibling technique.
15 . The system of claim 12 , wherein synthesizing the merged knowledge representation comprises using a substitution operation, wherein the substitution operation comprises using a retrieval operation or an addition operation.
16 . The system of claim 12 , wherein computing the semantic relevance score for a concept or attribute in the plurality of concepts and attributes semantically relevant to the active concept comprises using at least one measure of relevance from among generation certainty, concept productivity, Jaccard, statistical coherence, and cosine similarity.
17 . The system of claim 12 , wherein identifying or generating the active concept comprises:
determining whether the user context information matches an identifier of a concept in the merged knowledge representation; and
when it is determined that the user context information does not match the identifier of the concept in the merged knowledge representation, generating the active concept in the merged knowledge representation.
18 . The system of claim 12 , wherein the user context information comprises at least one of demographic information associated with the user, information from the user's browsing history, information typed in by the user, and information highlighted by the user.
19 . The system of claim 12 , wherein the target knowledge representation is represented by a data structure embodying a directed graph comprising a plurality of nodes and a plurality of edges, wherein each node is associated with a concept and an edge between two nodes represents a relationship between the two corresponding concepts.
20 . The system of claim 12 , wherein the target set of content comprises content accessible through an e-commerce website, a website of a business, a website providing access to one or more databases, an online portal, or a corporate Intranet.
21 . The system of claim 12 , wherein the reference set of digital content comprises content accessible through an information repository, database, content-provisioning source, or reference corpora.
22 . The system of claim 12 , wherein the user context information comprises a user-specified request or query.
23 . At least one non-transitory computer-readable storage medium storing processor-executable instructions that, when executed by at least one processor of at least one server computer, cause the at least one processor to perform a method of improving querying techniques used in providing to a search service textual information that is personalized to at least one user, the method comprising:
receiving user context information including textual information indicative of the interests associated with the at least one user;
identifying or generating, using a semantic analysis module, an active concept in a semantic network encoded in non-transitory memory, the active concept representing at least a portion of the user context information;
augmenting the user context information using a semantic analysis operations comprising:
retrieving one or more concepts from a target knowledge representation, the target knowledge representation and the reference knowledge representation encoded in non-transitory computer-readable memory, wherein the target knowledge representation represents knowledge from a target set of digital content indexed by a search engine, and wherein the reference knowledge representation represents a reference set of digital content;
analyzing the one or more concepts from the target knowledge representation and the one or more concepts from the reference knowledge representation to identify sets of target-attributes and reference-attributes, wherein the target-attributes and reference-attributes are more fundamental knowledge representation entities that comprise the one or more concepts from the target knowledge representation and the one or more concepts from the reference knowledge representation, respectively;
synthesizing a merged knowledge representation derived from merging the target knowledge representation and the reference knowledge representation based on the sets of target-attributes and reference-attributes around one or more concepts in common between the target knowledge representation and the reference knowledge representation, the knowledge representation encoded in non-transitory memory and comprising a plurality of concepts and attributes semantically relevant to the active concept;
determining a semantic relevance score for the plurality of concepts and attributes semantically relevant to the active concept using a measure of relevance, the semantic relevance score for each concept or each attribute of the plurality of concepts and attributes being indicative of semantic relevance of that concept or attribute to the active concept, and wherein the semantic relevance score is determined based at least in part on the structure of the merged knowledge representation; and
constructing one or more search queries derived from the user context information and labels of the plurality of concepts and attributes semantically relevant to the active concept, based on the semantic relevance score computed for each concept or attribute of the plurality of concepts and attributes semantically relevant to the active concept; and
returning the one or more search queries to the search service to retrieve textual information personalized to the at least one user.
24 . The at least one non-transitory computer-readable storage medium of claim 23 , wherein synthesizing the merged knowledge representation comprises pruning one or more concepts from the reference knowledge representation and/or from the target knowledge representation prior to the merging.
25 . The at least one non-transitory computer-readable storage medium of claim 23 , wherein synthesizing the merged knowledge representation comprises using an addition operation based on an analogy-by-parent technique and/or an analogy-by-sibling technique.
26 . The at least one non-transitory computer-readable storage medium of claim 23 , wherein synthesizing the merged knowledge representation comprises using a substitution operation, wherein the substitution operation comprises using a retrieval operation or an addition operation.
27 . The at least one non-transitory computer-readable storage medium of claim 23 , wherein computing the semantic relevance score for a concept or attribute in the plurality of concepts and attributes semantically relevant to the active concept comprises using at least one measure of relevance from among generation certainty, concept productivity, Jaccard, statistical coherence, and cosine similarity.
28 . The at least one non-transitory computer-readable storage medium of claim 23 , wherein identifying or generating the active concept comprises:
determining whether the user context information matches an identifier of a concept in the merged knowledge representation; and
when it is determined that the user context information does not match the identifier of the concept in the merged knowledge representation, generating the active concept in the merged knowledge representation.
29 . The at least one non-transitory computer-readable storage medium of claim 23 , wherein the user context information comprises at least one of demographic information associated with the user, information from the user's browsing history, information typed in by the user, and information highlighted by the user.
30 . The at least one non-transitory computer-readable storage medium of claim 23 , wherein the target knowledge representation is represented by a data structure embodying a directed graph comprising a plurality of nodes and a plurality of edges, wherein each node is associated with a concept and an edge between two nodes represents a relationship between the two corresponding concepts.
31 . The at least one non-transitory computer-readable storage medium of claim 23 , wherein the target set of content comprises content accessible through an e-commerce website, a website of a business, a website providing access to one or more databases, an online portal, or a corporate Intranet.
32 . The at least one non-transitory computer-readable storage medium of claim 23 , wherein the reference set of digital content comprises content accessible through an information repository, database, content-provisioning source, or reference corpora.
33 . The at least one non-transitory computer-readable storage medium of claim 23 , wherein the user context information comprises a user-specified request or query.