초록 ▼

A system filters documents in a document set retrieved from a document store in response to a query. The system obtains a first set of logical forms based on a selected one of the query and the documents in the document set. The system obtains a second set of logical forms based on another of the qu

A system filters documents in a document set retrieved from a document store in response to a query. The system obtains a first set of logical forms based on a selected one of the query and the documents in the document set. The system obtains a second set of logical forms based on another of the query and the documents in the document set. The system then uses natural language processing techniques to modify the first logical forms to obtain a modified set of logical forms. The system filters documents in the document set based on a predetermined relationship between the modified set of logical forms and the second set of logical forms.

대표청구항 ▼

1. A method of determining similar ity between first and second textual inputs, the method comprising:obtaining a first set of logical forms based on the first textual input; obtaining a second set of logical forms based on the second textual input; comparing the first and second sets of logical for

1. A method of determining similar ity between first and second textual inputs, the method comprising:obtaining a first set of logical forms based on the first textual input; obtaining a second set of logical forms based on the second textual input; comparing the first and second sets of logical forms; and determining similarity between the first and second textual inputs based on the step of comparing. 2. The method of claim 1 wherein comparing comprises:determining whether any logical forms in the first set match any logical forms in the second set. 3. The method of claim 2 wherein determining similarity comprises:assigning a score to reflect a degree of similarity between the first and second textual inputs based on matches between the first and second sets of logical forms. 4. The method of claim 1 and further comprising: obtaining a first set of paraphrased logical forms based on the first set of logical forms.5. The method of claim 4 wherein the comparing step comprises:comparing the first set of paraphrased logical forms with the second set of logical forms; and determining whether any paraphrased logical forms in the first set of paraphrased logical forms matches any logical forms in the second set of logical forms. 6. The method of claim 5 and further comprising:obtaining a second set of paraphrased logical forms based on the second set of logical forms. 7. The method of claim 6 wherein the comparing step further comprises:comparing the first set of paraphrased logical forms with the second set of paraphrased logical forms; and determining whether any paraphrased logical forms in the first set of paraphrased logical forms matches any paraphrased logical forms in the second set of paraphrased logical forms. 8. The method of claim 1 wherein the first textual input comprises an information retrieval query and wherein the second textual input comprises at least one document retrieved based on the query.9. The method of claim 1 wherein the second textual input comprises an information retrieval query and wherein the first textual input comprises at least one document retrieved based on the query.10. The method of claim 5, wherein the original structural relation comprises an original structural relation between the original words, and wherein obtaining a first set of paraphrased logical forms comprises:obtaining additional logical forms including expansion words, semantically related to the original words, and connected by the original structural relation. 11. The method of claim 9 wherein the original words include a first original word and a second original word connected by the original structural relation and wherein obtaining additional logical forms comprises at least one of:lexically expanding the first original word to include first related words which are semantically related to the first original word; lexically expanding the second original word to include second related words which are semantically related to the second original word; and connecting different ones of the first and second related words to one another by the original structural relation to obtain the additional logical forms. 12. The method of claim 11 wherein lexically expanding the first original word or lexically expanding the second original word comprises:obtaining synonyms for the first and second original words. 13. The method of claim 11 wherein lexically expanding the first original word or lexically expanding the second original word comprises:obtaining hypernyms for the first and second original words. 14. The method of claim 11 wherein lexically expanding the first original word or lexically expanding the second original word comprises:obtaining hyponyms for the first and second original words. 15. The method of claim 12 wherein obtaining a first set of paraphrased logical forms comprises:obtaining expanded structural relations related to the original structural relation; and connecting the original words with the expanded structural relations to obtain the paraphrased logical forms. 16. The method of claim 15 wherein obtaining a first set of logical forms further comprises:obtaining expansion words, semantically related to the original words; and connecting the expansion words with the original structural relation. 17. The method of claim 16 wherein obtaining a first set of paraphrased logical forms further comprises:connecting the expansion words with the expanded structural relations. 18. The method of claim 17 wherein the first set of logical forms includes at least one content word and wherein obtaining a first set of paraphrased logical forms comprises:obtaining a first set of abstract logical forms based on the content word. 19. The method of claim 18 wherein the first textual input comprises a document retrieval query and wherein obtaining a first set of abstract logical forms comprises:prior to generating the first set of abstract logical forms, identifying the query as a keyword query, in which the content word is unmodified by another content word, based on a structure of the query. 20. The method of claim 19 wherein the second textual input comprises a document, and further comprising:obtaining a second set of paraphrased logical forms based on the second set of logical forms. 21. The method of claim 20 wherein obtaining the second set of logical forms comprises:obtaining a set of abstract logical forms indicative of a meta structure of the document. 22. The method of claim 21 wherein the meta structure of the document is indicative of a general subject matter of the document.23. The method of claim 21, wherein obtaining the set of abstract logical forms indicative of a meta structure of the document comprises:obtaining the set of abstract logical forms based on formatting information corresponding to the document. 24. The method of claim 22 wherein obtaining the set of abstract logical forms indicative of a meta structure of the document comprises:obtaining the set of abstract logical forms based on topics of sentences in the document. 25. The method of claim 22 wherein obtaining the set of abstract logical forms indicative of a meta structure of the document comprises:obtaining the set of abstract logical forms based on subjects of sentences in the document. 26. The method of claim 20 and further comprising:suppressing other logical forms based on the content word, other than the first and second set of paraphrased logical forms. 27. A method of filtering documents in a document set retrieved from a document store in response to a query, the method comprising:obtaining a first set of logical forms based on a selected one of the query and the documents in the document set; obtaining a second set of logical forms based on another of the query and the documents in the document set; obtaining a first set of paraphrased logical forms indicative of paraphrases of at least the first set of logical forms; and filtering documents in the document set based on a predetermined relationship between the first set of paraphrased logical forms and the second set of logical forms. 28. The method of claim 27 wherein obtaining the first set of logical forms comprises:obtaining original words and an original structural relation between the original words based on the selected one of the query and the documents in the document set. 29. The method of claim 28 wherein the original structural relation comprises an original structural relation between the original words, and wherein obtaining a first set of paraphrased logical forms comprises:obtaining additional logical forms including expansion words, semantically related to the original words, and connected by the original structural relation. 30. The method of claim 29 wherein the original words include a first original word and a second original word connected by the original structural relation and wherein obtaining additional logical forms comprises at least one of:lexically expanding the first original word to include first related words which are semantically related to the first original word; lexically expanding the second original word to include second related words which are semantically related to the second original word; and connecting different ones of the first and second related words to one another by the original structural relation to obtain the additional logical forms. 31. The method of claim 30 wherein lexically expanding the first original word and lexically expanding the second original word comprises:obtaining synonyms for the first or second original words. 32. The method of claim 30 wherein lexically expanding the first original word and lexically expanding the second original word comprises:obtaining hypernyms for the first or second original words. 33. The method of claim 30 wherein lexically expanding the first original word and lexically expanding the second original word comprises:obtaining hyponyms for the first or second original words. 34. The method of claim 28 wherein obtaining a first set of paraphrased logical forms comprises:obtaining expanded structural relations syntactically related to the original structural relation; and connecting the original words with the expanded structural relations to obtain the paraphrased logical forms. 35. The method of claim 34 wherein obtaining a first set of logical forms further comprises:obtaining expansion words, semantically related to the original words; and connecting the expansion words with the original structural relation. 36. The method of claim 35 wherein obtaining a first set of paraphrased logical forms further comprises:connecting the expansion words with the expanded structural relations. 37. The method of claim 27 wherein the first set of logical forms includes at least one content word and wherein obtaining a first set of paraphrased logical forms comprises:obtaining a first set of abstract logical forms based on the content word. 38. The method of claim 37 wherein the selected one of the query and the documents in the document set comprises the query and wherein obtaining a first set of abstract logical forms comprises:prior to generating the first set of abstract logical forms, identifying the query as a keyword query, in which the content word is unmodified by another content word, based on a structure of the query. 39. The method of claim 38 wherein the second set of logical forms is based on the documents in the document set, and further comprising:obtaining a second set of paraphrased logical forms based on the second set of logical forms. 40. The method of claim 39 wherein obtaining the second set of logical forms comprises:obtaining a set of abstract logical forms indicative of a meta structure of each of the documents in the document set. 41. The method of claim 40 wherein the meta structure of each document is indicative of a general subject matter of the document.42. The method of claim 41 wherein obtaining the set of abstract logical forms indicative of a meta structure of each of the documents comprises:obtaining the set of abstract logical forms based on formatting information corresponding to each document. 43. The method of claim 41 wherein obtaining the set of abstract logical forms indicative of a meta structure of each of the documents comprises:obtaining the set of abstract logical forms based on topics of sentences in each document. 44. The method of claim 41 wherein obtaining the set of abstract logical forms indicative of a meta structure of each of the documents comprises:obtaining the set of abstract logical forms based on subjects of sentences in each document. 45. The method of claim 39 and further comprising:suppressing other logical forms based on the content word, other than the first and second set of paraphrased logical forms. 46. The method of claim 27, wherein filtering comprises:providing an output indicative of a ranked order of the documents in the document set based on the predetermined relationship. 47. A method of filtering documents in a document set retrieved from a document store in response to a query, the method comprising:obtaining a first set of logical forms based on a selected one of the query and the document set; obtaining a second set of logical forms based on based on another of the query and the document set; suppressing a first predetermined class of logical forms in at least the first set of logical forms to obtain a first suppressed set of logical forms; and filtering the documents in the document set based on a predetermined relationship between the first suppressed set of logical forms and the second set of logical forms. 48. The method of claim 47 wherein suppressing comprises:suppressing logical forms having a predetermined structure. 49. The method of claim 47 wherein suppressing comprises:suppressing logical forms occurring with a frequency which exceeds a threshold frequency level. 50. The method of claim 47 and further comprising:suppressing a second predetermined class of logical forms in the second set of logical forms, the second predetermined class being different from the first predetermined class. 51. The method of claim 47 wherein suppressing is performed prior to obtaining the first set of logical forms.52. The method of claim 47 wherein suppressing is performed substantially simultaneously with obtaining the first set of logical forms.53. The method of claim 47 wherein suppressing is performed after obtaining the first set of logical forms.54. A computer readable medium including computer readable data stored thereon, the computer readable data including:index data indicative of contents of documents in a document set; and a set of abstract logical forms indicative of a meta structure of each of the documents in the document set, the abstract logical forms representing information not found in an explicit expression in corresponding documents. 55. The computer readable medium of claim 54 wherein the meta structure of each document is indicative of a general subject matter of the document.56. The computer readable medium of claim 55 wherein the set of abstract logical forms are based on formatting information corresponding to each document.57. The computer readable medium of claim 55 wherein the abstract logical forms are based on topics of sentences in each document.58. The computer readable medium of claim 55 wherein the set of abstract logical forms are based on subjects of sentences in each document.59. A computer readable medium including computer readable instructions stored thereon which, when executed by the computer, cause the computer to filter documents in a document set from a document store in response to a query by performing the steps of:obtaining a first set of logical forms based on a selected one of the query and the documents in the document set; obtaining a second set of logical forms based on another of the query and the documents in the document set; using natural language processing to modify at least the first set of logical forms to obtain a first modified set of logical forms; and filtering documents in the document set based on a predetermined relationship between the first modified set of logical forms and the second set of logical forms. 60. The computer readable medium of claim 59 wherein using natural language processing comprises:obtaining a first set of paraphrased logical forms indicative of paraphrases of at least the first set of logical forms. 61. The computer readable medium of claim 60 wherein obtaining the first set of logical forms comprises:obtaining original words and an original structural relation between the original words based on the selected one of the query and the documents in the document set. 62. The computer readable medium of claim 61 wherein the original structural relation comprises an original structural relation between the original words, and wherein obtaining a first set of paraphrased logical forms comprises:obtaining additional logical forms including expansion words, semantically related to the original words, and connected by the original structural relation. 63. The computer readable medium of claim 61 wherein obtaining a first set of paraphrased logical forms comprises:obtaining expanded structural relations syntactically related to the original structural relation; and connecting the original words with the expanded structural relations to obtain the paraphrased logical forms. 64. The computer readable medium of claim 59 wherein the first set of logical forms includes at least one content word and wherein obtaining a first set of paraphrased logical forms comprises:obtaining a first set of abstract logical forms based on the content word. 65. The computer readable medium of claim 64 wherein the selected one of the query and the documents in the document set comprises the query and wherein obtaining a first set of abstract logical forms comprises:prior to generating the first set of abstract logical forms, identifying the query as a keyword query, in which the content word is unmodified by another content word, based on a structure of the query. 66. The computer readable medium of claim 59 wherein the first set of logical forms is based on the documents in the document set, and wherein using natural language processing comprises:obtaining a set of abstract logical forms indicative of a meta structure of each of the documents in the document set. 67. The computer readable medium of claim 59 wherein using natural language processing comprises:suppressing a first predetermined class of logical forms in at least the first set of logical forms to obtain a first suppressed set of logical forms. 68. The computer readable medium of claim 67 wherein suppressing comprises:suppressing logical forms having a predetermined structure. 69. The computer readable medium of claim 67 wherein suppressing comprises:suppressing logical forms occurring with a frequency which exceeds a threshold frequency level. 70. A method of determining similarity between first and second textual inputs, the method comprising:obtaining a first set of logical forms based on the first textual input; obtaining a second set of logical forms based on the second textual input; suppressing a first predetermined class of logical forms in at least the first set of logical forms to obtain a first suppressed set of logical forms; and determining similarity between the first and second textual inputs by comparing the first suppressed set of logical forms and the second set of logical forms. 71. The method of claim 70 wherein suppressing comprises:suppressing logical forms having a predetermined structure. 72. The method of claim 70 wherein suppressing comprises:suppressing logical forms occurring with a frequency which exceeds a threshold frequency level. 73. The method of claim 70 and further comprising:suppressing a second predetermined class of logical forms in the second set of logical forms, the second predetermined class being different from the first predetermined class. 74. The method of claim 70 wherein suppressing is performed prior to obtaining the first set of logical forms.75. The method of claim 70 wherein suppressing is performed substantially simultaneously with obtaining the first set of logical forms.76. The method of claim 70 wherein suppressing is performed after obtaining the first set of logical forms.