Identifying superphrases of text strings
Methods and apparatus are described by which “superphrases” of “seed phrases” representing basic concepts may be identified without having to compare all possible pairs of seed and candidate phrases. According to one class of embodiments, a data structure similar to an inverted index is used for indexing phrases. The elimination of seed and candidate phrase pairs is enabled by building and traversing the index in a particular manner.
1. A computer-implemented method for rewriting search queries with reference to a set of seed queries, comprising:
determining whether a first search query includes one or more of the seed queries;
where the first search query includes a first one of the seed queries, reformulating the first search query as a rewritten search query, wherein the rewritten search query includes the first seed query and specifies different matching requirements for corresponding subsets of terms included in the rewritten search query, each matching requirement imposing a rule governing how terms in the corresponding subset of terms must appear in or be associated with matching documents, and wherein the matching requirement corresponding to the first seed query is more restrictive than the matching requirements corresponding to any other subsets of terms of the rewritten search query; and
generating search results with reference to the rewritten search query and the different matching requirements in response to the first search query.
2. The method of claim 1 wherein a first one of the subsets of terms included in the rewritten search query corresponds to the first seed query.
3. The method of claim 2 wherein generating the search results comprises identifying documents having titles matching the first seed query, and additional text matching any remaining terms in the rewritten query beyond the first seed query.
4. A computer-implemented method for responding to search queries, comprising facilitating presentation of search results in an interface in response to a first search query, the search results having been generated with reference to a rewritten search query which corresponds to a reformulation of the first search query, the rewritten search query including a seed query contained within the first search query and specifying different matching requirements for corresponding subsets of terms included in the rewritten search query, each matching requirement imposing a rule governing how terms in the corresponding subset of terms must appear in or be associated with matching documents, and wherein the matching requirement corresponding to the seed query is more restrictive than the matching requirements corresponding to any other subsets of terms of the rewritten search query, the search results also having been generated with reference to the different matching requirements.
5. The method of claim 4 wherein the search results include references to documents having titles matching the first seed query, and additional text matching any remaining terms in the rewritten query beyond the first seed query.
6. A computer program product for rewriting search queries with reference to a set of seed queries, the computer program product comprising at least one non-transitory computer-readable medium having computer program instructions stored therein configured to enable at least one computing device to:
determine whether a first search query includes one or more of the seed queries;
where the first search query includes a first one of the seed queries, reformulate the first search query as a rewritten search query, wherein the rewritten search query includes the first seed query and specifies different matching requirements for corresponding subsets of terms included in the rewritten search query, each matching requirement imposing a rule governing how terms in the corresponding subset of terms must appear in or be associated with matching documents, and wherein the matching requirement corresponding to the first seed query is more restrictive than the matching requirements corresponding to any other subsets of terms of the rewritten search query; and
generate search results with reference to the rewritten search query and the different matching requirements in response to the first search query.
7. The computer program product of claim 6 wherein a first one of the subsets of terms included in the rewritten search query corresponds to the first seed query.
8. The computer program product of claim 7 wherein the computer program instructions are configured to enable the at least one computing device to generate the search results by identifying documents having titles matching the first seed query, and additional text matching any remaining terms in the rewritten query beyond the first seed query.
9. A system for rewriting search queries with reference to a set of seed queries, the system comprising at least one computing device comprising at least one processor coupled to at least one memory and configured to:
determine whether a first search query includes one or more of the seed queries;
where the first search query includes a first one of the seed queries, reformulate the first search query as a rewritten search query, wherein the rewritten search query includes the first seed query and specifies different matching requirements for corresponding subsets of terms included in the rewritten search query, each matching requirement imposing a rule governing how terms in the corresponding subset of terms must appear in or be associated with matching documents, and wherein the matching requirement corresponding to the first seed query is more restrictive than the matching requirements corresponding to any other subsets of terms of the rewritten search query; and
generate search results with reference to the rewritten search query and the different matching requirements in response to the first search query.