Recommendation system with hierarchical mapping and imperfect matching
Aspects described herein provide solutions to problems posed by a user. Input that includes a specified subject may be received from a user. A specified descriptor for the specified subject may be obtained from a hierarchical taxonomy associated with the specified subject. An information repository may be searched based on the specific descriptor, and entries stored at the information repository that are associated with the specific descriptor may be indicated in a list of results. The specific descriptor may be iteratively generalized to obtain generalized descriptors, and the information repository may be searched based on the generalized descriptors. The generalized descriptors may also be specialized to obtain specialized descriptors, and the information repository may be searched based on the specialized descriptors. The list of results may include entries stored at the information repository that are respectively associated with the generalized descriptors and the specialized descriptors.
1. A computer-implemented method, comprising:
receiving input from a user wherein the input comprises a specified subject;
obtaining a specific descriptor for the specified subject from a hierarchical taxonomy associated with the specified subject;
searching an information repository based on the specific descriptor and indicating in results one or more entries stored at the information repository that are associated with the specific descriptor;
iteratively generalizing the specific descriptor to obtain at least one generalized descriptor, wherein generalizing the specific descriptor includes removing a level from the specific descriptor for the hierarchical taxonomy associated with the specific descriptor;
searching the information repository based on one or more of the generalized descriptors and indicating in the results one or more entries stored at the information repository that are associated with one of the generalized descriptors;
iteratively specializing one or more of the generalized descriptors to obtain at least one specialized descriptor, wherein specializing the one or more of the generalized descriptors includes adding a level from one of the one or more generalized descriptors for the hierarchical taxonomy associated with the specific descriptor; and
searching the information repository based on one or more of the specialized descriptors and indicating in the results one or more entries stored at the information repository that are associated with one of the specialized descriptors.
2. The computer-implemented method of claim 1 wherein the hierarchical taxonomy is a first hierarchical taxonomy and further comprising:
searching a taxonomy mapping that maps at least one descriptor associated with the first hierarchical taxonomy to at least one related descriptor of a second hierarchical taxonomy;
identifying a mapping between the specific descriptor and one of the related descriptors of the taxonomy mapping; and
wherein searching the information repository based on the specific descriptor comprises searching the information repository using the related descriptor.
3. The computer-implemented method of claim 2 wherein:
searching the information repository based on one or more of the generalized descriptors comprises searching the information repository using one of the related descriptors that is mapped to one of the generalized descriptors; and
searching the information repository based on one or more of the specialized descriptors comprises searching the information repository using one of the related descriptors that is mapped to one of the specialized descriptors.
4. The computer-implemented method of claim 2 further comprising:
iteratively specializing one of the related descriptors to obtain at least one specialized related descriptor; and
searching the information repository using one of the specialized related descriptors and indicating in the results one or more entries stored at the information repository that are associated with the specialized related descriptor.
5. The computer-implemented method of claim 2 further comprising:
scoring individual results in the results to obtain respective result scores for the individual results; and
ordering the results based on the respective result scores.
6. The computer-implemented method of claim 5 wherein one of the result scores for one of the results in the results is based, at least in part, on a number of generalizations or a number of specializations performed in order to locate the result in the information repository.
7. The computer-implemented method of claim 6 wherein:
the result is a place-of-interest having a geographic location; and
the result score for the place-of-interest is further based on a distance between the geographic location of the place-of-interest and a current location of the user.
8. The computer-implemented method of claim 2 wherein:
the specified subject is a specified product the user desires to purchase;
the first taxonomy is a product taxonomy of products and the specific descriptor is a specific product descriptor;
the second taxonomy is a place-of-interest taxonomy of places-of-interest and the related descriptor is a place-of-interest descriptor corresponding to a place-of-interest that provides the specified product; and
the results include one or more places-of-interest.
9. The computer-implemented method of claim 8 further comprising including geographic information in the results wherein the geographic information indicates respective geographic locations of one or more of the places-of-interest in the results.
10. The computer-implemented method of claim 1 wherein:
the specific descriptor comprises a plurality of classifications that describe the specified subject and that are nested relative to one another;
generalizing the specific descriptor comprises removing the most specific classification from the specific descriptor; and
specializing one of the generalized descriptors comprises adding a new classification to the generalized descriptor that is more specific relative to each classification of the generalized descriptor.
11. A system for providing possible solutions to a problem posed by a user comprising:
at least one processing unit;
memory storing computer-readable instructions that, when executed by the processing unit, cause the system to provide:
an interface configured to receive input from a user, where the input comprises a specified subject; and
a recommendation engine configured to obtain a specific descriptor for the specified subject from a hierarchical taxonomy associated with the specified subject, iteratively generalize the specific descriptor to obtain at least one generalized descriptor, wherein generalizing the specific descriptor includes removing a level from the specific descriptor for the hierarchical taxonomy associated with the specific descriptor, iteratively specialize one or more of the generalized descriptors to obtain at least one specialized descriptor, wherein specializing the one or more of the generalized descriptors includes adding a level from one of the one or more generalized descriptors for the hierarchical taxonomy associated with the specific descriptor, and search an information repository based on the specific descriptor, the one or more generalized descriptors, and the one or more specialized descriptors to obtain results;
wherein the results comprises indications of entries stored at the information repository that are respectively associated with at least one of the specific descriptor, the one or more generalized descriptors, and the one or more specialized descriptors.
12. The system of claim 11 wherein:
the hierarchical taxonomy is a source hierarchical taxonomy; and
the recommendation engine is further configured to search a taxonomy mapping that maps at least one descriptor associated with the first hierarchical taxonomy to at least one related descriptor of a target hierarchical taxonomy;
identify a mapping between the specific descriptor and one of the related descriptors of the taxonomy mapping; and
search the information repository using the related descriptor.
13. The system of claim 12 wherein the recommendation engine is further configured to search the information repository using one of the related descriptors that is mapped to one of the generalized descriptors and search the information repository using one of the related descriptors that is mapped to one of the specialized descriptors.
14. The system of claim 12 wherein:
the recommendation engine is further configured to iteratively specialize one of the related descriptors to obtain at least one specialized related descriptor and search the information repository using one of the specialized related descriptors; and
the results further comprises indications of one or more entries stored at the information repository that are associated with the specialized related descriptor.
15. The system of claim 12 wherein:
the computer-readable instructions, when executed by the at least one processing unit, further cause the system to provide a scoring module; and
the scoring module is configured to score individual results in the results to obtain respective result scores for the individual results and order the results based on the respective result scores.
16. The system of claim 15 wherein one of the result scores for one of the results in the results is based, at least in part, on a number of generalizations or a number of specializations performed in order to locate the result in the information repository.
17. The system of claim 16 wherein:
the result is a place-of-interest having a geographic location; and
the result score for the place-of-interest is further based on a distance between the geographic location of the place-of-interest and a current location of the user.
18. The system of claim 12 wherein:
the specified subject is a specified product the user desires to purchase;
the source taxonomy is a product taxonomy of products and the specific descriptor is a specific product descriptor;
the target taxonomy is a place-of-interest taxonomy of places-of-interest and the related descriptor is a place-of-interest descriptor corresponding to a place-of-interest that provides the specified product; and
the results include one or more places-of-interest.
19. The system of claim 18 wherein the recommendation engine is further configured to include geographic information in the results wherein the geographic information indicates respective geographic locations of one or more of the places-of-interest in the results.
20. The system of claim 11 wherein:
the specific descriptor comprises a plurality of classifications that describe the specified subject and that are nested relative to one another;
the recommendation engine is configured to generalize the specific descriptor by removing the most specific classification from the specific descriptor; and
the recommendation engine is configured to specialize one of the generalized descriptors by adding a new classification to the generalized descriptor that is more specific relative to each classification of the generalized descriptor.