Search engine method and apparatus

• US 20030217052A1
• Filed: 05/14/2003
• Published: 11/20/2003
• Est. Priority Date: 08/24/2000
• Status: Abandoned Application

• Current Assignee: Celebros Ltd.
• Original Assignee: Celebros Ltd.
• Inventors: Roth, Assaf, Flor, Michael, Dershowitz, Nachum, Choueka, Yaacov, Hod, Oren, Rubenczyk, Tal
• Application Number:
  US10/436,996
• Publication Number:
  US 20030217052A1
• Time in Patent Office: Days
• US Class Current: 707/3
• CPC Class Codes: G06F 16/3323 using document space presentation or visualization, e.g. category, hierarchy or range presentation and selection
• CPC Class Codes: G06F 16/3325 Reformulation based on results of preceding query
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06N 20/00 Machine learning

Abstract

An interactive method for searching a database to produce a refined results space, the method comprising: analyzing for search criteria, searching said database using said search criteria to obtain an initial result space, and obtaining user input to restrict said initial results space, thereby to obtain said refined results space. Refining comprises using classifications of the retrieved data items to formulate prompts for the user, asking said user at least one of the formulated prompts and receiving a response thereto; and using responses in conjunction with classification values to exclude some of the results, thereby to provide to the user a subset of the retrieved data items as a query result.

Claims

1. An interactive method for searching a database to produce a refined results space, the method comprising:

analyzing for search criteria,

searching said database using said search criteria to obtain an initial result space, and

obtaining user input to restrict said initial results space, thereby to obtain said refined results space.

2. The method of claim 1, wherein said searching comprises browsing.

3. The method of claim 1, wherein said analyzing is performed on said database prior to searching, thereby to optimize said database for said searching.

4. The method of claim 1, wherein said analyzing is performed on a search criterion input by a user.

5. The method of claim 1, wherein said analyzing comprises using linguistic analysis.

6. The method of claim 4, comprising carrying out said analyzing on an initial search criterion to obtain an additional search criterion.

7. The method of claim 6, wherein said search criterion is a null criterion.

8. The method of claim 6, wherein said analyzing for additional search criteria is carried out using linguistic analysis of said initial search criterion.

9. The method of claim 1, wherein said analyzing is carried out by selection of related concepts.

10. The method of claim 1, wherein said analyzing is carried out using data obtained from past operation of said method.

11. The method of claim 1, comprising generating a prompt for said obtaining user input, by generating at least one prompt having at least two answers, said answers being selected to divide said initial results space.

12. The method of claim 11, wherein said generating a prompt comprises generating at least one segmenting prompt having a plurality of potential answers, each answer corresponding to a part of said results space.

13. The method of claim 12, wherein each part of said results space comprises a substantially proportionate share of said results space.

14. The method of claim 12, comprising generating a plurality of segmenting prompts and choosing therefrom a prompt whose answers most evenly divide said results space.

15. The method of claim 11, wherein said restricting said results space comprises rejecting, from said results space, any results not corresponding to an answer given in said user input.

16. The method of claim 15, further comprising allowing a user to insert additional text, said text being usable as part of said user input in said restricting.

17. The method of claim 11, further comprising repeating said obtaining user input by generating at least one further prompt having at least two answers, said answers being selected to divide said refined results space.

18. The method of claim 17, comprising continuing said restricting until said refined results space is contracted to a predetermined size.

19. The method of claim 17, comprising continuing said restricting until no further prompts are found.

20. The method of claim 17, comprising continuing said restricting until a user input is received to stop further restriction and submit the existing results space.

21. The method of claim 17, further comprising determining that a submitted results space does not include a desired item, and following said determination to submit to said user initially retrieved items that have been excluded by said restricting.

22. The method of claim 20, further comprising:

obtaining from a user a determination that a submitted results space does not include a desired item, and

submitting to said user initially retrieved items that have been excluded by said restricting.

23. The method of claim 1, comprising receiving said initial search criterion as user input.

24. The method of claim 11, wherein said obtaining said user input includes providing a possibility for a user not to select an answer to said prompt.

25. The method of claim 24, further comprising asking an additional prompt following non-selection of an answer by said user.

26. The method of claim 1, further comprising updating system internal search-supporting information according to a final selection of an item by a user following a query.

27. The method of claim 26, wherein said updating comprises modifying a correlation between said selected item and said obtained user input.

28. Apparatus for interactively searching a database to produce a refined results space, comprising:

a search criterion analyzer for analyzing to obtain search criteria,

a database searcher, associated with said search criterion analyzer, for searching said database using said search criteria to obtain an initial result space, and

a restrictor, for obtaining user input to restrict said results space, and using said user input to restrict said results space, thereby to formulate a refined results space.

29. The apparatus of claim 28, wherein said search criterion analyzer comprises a database data-items analyzer capable of producing classifications for data items to correspond with analyzed search criteria.

30. The apparatus of claim 28, wherein said search criterion analyzer comprises a database data-items analyzer capable of utilizing classifications for data items to correspond with analyzed search criteria.

31. The apparatus of claim 29, wherein said search criterion analyzer is further capable of utilizing classifications for data items to correspond with analyzed search criteria.

32. The apparatus of claim 29, wherein said database data items analyzer is operable to analyze at least part of said database prior to said search.

33. The apparatus of claim 29, wherein said database data items analyzer is operable to analyze at least part of said database during said search.

34. The apparatus of claim 28, wherein said analyzing comprises linguistic analysis.

35. The apparatus of claim 28, wherein said analyzing comprises statistical analysis.

36. The apparatus of claim 34, wherein said analyzing comprises statistical language-analysis.

37. The apparatus of claim 28, wherein said search criterion analyzer is configured to receive an initial search criterion from a user for said analyzing.

38. The apparatus of claim 37, wherein said initial search criterion is a null criterion.

39. The apparatus of claim 37, wherein said analyzer is configured to carry out linguistic analysis of said initial search criterion.

40. The apparatus of claim 28, wherein said analyzer is configured to carry out an analysis based on selection of related concepts.

41. The apparatus of claim 28, wherein said analyzer is configured to carry out an analysis based on historical knowledge obtained over previous searches.

42. The apparatus of claim 28, wherein said restrictor is operable to generate a prompt for said obtaining user input, said prompt comprising at least two selectable responses, said responses being usable to divide said initial results space.

43. The apparatus of claim 42, wherein said prompt comprises a segmenting prompt having a plurality of potential answers, each answer corresponding to a part of said results space, and each part comprising a substantially proportionate share of said results space.

44. The apparatus of claim 42, wherein generating said prompt comprises

generating a plurality of segmenting prompts, each having a plurality of potential answers, each answer corresponding to a part of said results space, and each part comprising a substantially proportionate share of said results space, and

selecting one of said prompts whose answers most evenly divide said results space.

45. The apparatus of claim 42, further comprising allowing a user to insert additional text, said text being usable as part of said user input by said restrictor.

46. The apparatus of claim 42, wherein said restricting said results space comprises rejecting therefrom any results not corresponding to an answer given in said user input, thereby to generate a revised results space.

47. The apparatus of claim 46, wherein said restrictor is operable to generate at least one further prompt having at least two answers, said answers being selected to divide said revised results space.

48. The apparatus of claim 47, wherein said restrictor is configured to continue said restricting until said refined results space is contracted to a predetermined size.

49. The apparatus of claim 47, wherein said restrictor is configured to continue said restricting until no further prompts are found.

50. The apparatus of claim 47, wherein said restrictor is configured to continue said restricting until a user input is received to stop further restriction and submit the existing results space.

51. The apparatus of claim 50, wherein a user is enabled to respond that a submitted results space does not include a desired item, the apparatus being configured to submit to said user initially retrieved items that have been excluded by said restricting, in receipt of such a response.

52. The apparatus of claim 47, comprising operability to determine that a submitted results space does not include a desired item, the apparatus being configured, following such a determination, to submit to said user initially retrieved items that have been excluded by said restricting, in receipt of such a response.

53. The apparatus of claim 28, wherein said analyzer is configured to receive said initial search criterion as user input.

54. The apparatus of claim 42, wherein said restrictor is configured to provide, with said prompt, a possibility for a user not to select an answer to said prompt.

55. The apparatus of claim 54, wherein said restrictor is operable to provide a further prompt following non-selection of an answer by said user.

56. The apparatus of claim 28, further comprising an updating unit for updating system internal search-supporting information according to a final selection of an item by a user following a query.

57. The apparatus of claim 56, wherein said updating comprises modifying a correlation between said selected item and said obtained user input.

58. The apparatus of claim 56, wherein said updating comprises modifying a correlation between a classification of said selected item and said obtained user input.

59. A database with apparatus for interactive searching thereof to produce a refined results space, the apparatus comprising:

a search criterion analyzer for analyzing for search criteria,

a database searcher, associated with said search criterion analyzer, for searching said database using search criteria to obtain an initial result space, and

a restrictor, for obtaining user input to restrict said results space, and using said user input to restrict said results space, thereby to provide said refined results space.

60. The apparatus of claim 59, wherein said search criterion analyzer comprises a database data-items analyzer capable of producing classifications for data items to correspond with analyzed search criteria.

61. The database of claim 59, wherein said search criterion analyzer comprises a database data-items analyzer capable of utilizing classifications for data items to correspond with analyzed search criteria.

62. The database of claim 60, wherein said database data items analyzer is further capable of utilizing classifications for data items to correspond with analyzed search criteria.

63. The database of claim 59, wherein said search criterion analyzer comprises a search criterion analyzer capable of analyzing user-provided search criteria in terms of a classification structure of items in said database.

64. The database of claim 59, comprising data items and wherein each data item is analyzed into potential search criteria, thereby to optimize matching with user input search criteria.

65. The database of claim 60, wherein said database data items analyzer is operable to carry out linguistic analysis.

66. The database of claim 60, wherein said database data items analyzer is operable to carry out statistical analysis.

67. The database of claim 65, wherein said database data items analyzer is operable to carry out statistical analysis.

68. The database of claim 59, wherein said search criterion analyzer is configured to receive an initial search criterion from a user for said analyzing.

69. The database of claim 68, wherein said initial search criterion is a null criterion.

70. The database of claim 68, wherein said analyzer is configured to carry out linguistic analysis of said initial search criterion.

71. The database of claim 59, wherein said analyzer is configured to carry out an analysis based on selection of related concepts.

72. The database of claim 59, wherein said analyzer is configured to carry out an analysis based on historical knowledge obtained over previous searches.

73. The database of claim 59, wherein said restrictor is operable to generate a prompt for said obtaining user input, said prompt comprising a prompt having at least two answers, said answers being selected to divide said initial results space.

74. The database of claim 73, wherein said prompt is a segmenting prompt having a plurality of potential answers, each answer corresponding to a part of said results space, and each part comprising a substantially proportionate share of said results space.

75. The database of claim 59, further comprising allowing a user to insert additional text, said text being usable as part of said user input by said restrictor.

76. The database of claim 73, wherein said restricting said results space comprises rejecting therefrom any results not corresponding to one of said answers of said user input, thereby to generate a revised results space.

77. The database of claim 76, wherein said restrictor is operable to generate at least one further prompt having at least two answers, said answers being selected to divide said revised results space.

78. The database of claim 77, wherein said restrictor is configured to continue said restricting until said refined results space is contracted to a predetermined size.

79. The database of claim 77, wherein said restrictor is configured to continue said restricting until no further prompts are found.

80. The database of claim 77, wherein said restrictor is configured to continue said restricting until a user input is received to stop further restriction and submit the existing results space.

81. The database of claim 80, wherein said user is enabled to respond that a submitted results space does not include a desired item, the database being operable in receipt of such a response to submit to said user initially retrieved items that have been excluded by said restricting.

82. The database of claim 77, further being operable to determine that a submitted results space does not include a desired item, the database being operable following such a determination to submit to said user initially retrieved items that have been excluded by said restricting.

83. The database of claim 59, wherein said analyzer is configured to receive said initial search criterion as user input.

84. The database of claim 73, wherein said restrictor is configured to provide, with said prompt, a possibility for a user not to select an answer to said prompt.

85. The database of claim 84, wherein said restrictor is further configured to provide an additional prompt following non-selection of an answer by said user.

86. The database of claim 59, further comprising an updating unit for updating system internal search-supporting information according to a final selection of an item by a user following a query.

87. The database of claim 86, wherein said updating comprises modifying a correlation between said selected item and said obtained user input.

88. The database of claim 86, wherein said updating comprises modifying a correlation between a classification of said selected item and said obtained user input.

89. A query method for searching stored data items, the method comprising:

i) receiving a query comprising at least a first search term,

ii) expanding the query by adding to said query, terms related to said at least first search term,

iii) retrieving data items corresponding to at least one of said terms,

iv) using attribute values applied to said retrieved data items to formulate prompts for said user,

v) asking said user at least one of said formulated prompts as a prompt for focusing said query,

vi) receiving a response thereto, and

vii) using said received response to compare to values of said attributes to exclude ones of said retrieved items, thereby to provide a subset of said retrieved data items as a query result.

90. The method of claim 89, wherein said query comprises a plurality of terms, and said expanding said query further comprises analyzing said terms to determine a grammatical interrelationship between ones of said terms.

91. The method of claim 90, further comprising using said grammatical interrelationship to identify leading and subsidiary terms of said search query.

92. The method of claim 89, wherein said expanding comprises a three-stage process of separately adding to said query:

a) items which are closely related to said search term,

b) items which are related to said search term to a lesser degree and

c) an alternative interpretation due to any ambiguity inherent in said search term.

93. The method of claim 92, wherein said items are one of a group comprising lexical terms and conceptual representations.

94. The method of claim 89, further comprising at least one additional focusing process of repeating stages iii) to vi), thereby to provide refined subsets of said retrieved data items as said query result.

95. The method of claim 89, further comprising ordering said formulated prompts according to an entropy weighting based on probability values and asking ones of said prompts having more extreme entropy weightings.

96. The method of claim 95, further comprising recalculating said probability values and consequently said entropy weightings following receiving of a response to an earlier prompt.

97. The method of claim 95, further comprising using a dynamic answer set for each prompt, said dynamic answer set comprising answers associated with classification values, said classification values being true for some received items and false for other received items, thereby to discriminate between said retrieved items.

98. The method of claim 97, further comprising ranking respective answers within said dynamic answer set according to a respective power to discriminate between said retrieved items.

99. The method of claim 95, further comprising modifying said probability values according to user search behavior.

100. The method of claim 99, wherein said user search behavior comprises past behavior of a current user.

101. The method of claim 99, wherein said user search behavior comprises past behavior aggregated over a group of users.

102. The method of claim 99, wherein said modifying comprises using said user search behavior to obtain a priori selection probabilities of respective data items, and modifying said weightings to reflect said probabilities.

103. The method of claim 95, wherein said entropy weighting is associated with at least one of a group comprising said items classifications of said items and respective classification values.

104. The method of claim 89, comprising semantically analyzing said stored data items prior to said receiving a query.

105. The method of claim 89, comprising semantically analyzing said stored data items during a search session.

106. The method of claim 104, wherein said semantic analysis comprises classifying said data items into classes.

107. The method of claim 106, further comprising classifying attributes into attribute classes.

108. The method of claim 106, wherein said classifying comprises distinguishing both among object-classes or major classes, and among attribute classes.

109. The method of claim 108, wherein said classifying comprises providing a plurality of classifications to a single data item.

110. The method of claim 106, wherein a classification arrangement of respective classes is pre-selected for intrinsic meaning to the subject-matter of a respective database.

111. The method of claim 110, comprising arranging major ones of said classes hierarchically.

112. The method of claim 107, comprising arranging attribute classes hierarchically.

113. The method of claim 112, further comprising determining semantic meaning for a term in said data item from a hierarchical arrangement of said term.

114. The method of claim 111, wherein said classes are also used in analyzing said query.

115. The method of claim 110, wherein attribute values are assigned weightings according to the subject-matter of a respective database.

116. The method of claim 110, wherein at least one of said attribute values and said classes are assigned roles in accordance with the subject-matter of a respective database.

117. The method of claim 116, wherein said roles are additionally used in parsing said query.

118. The method of claim 117, further comprising assigning importance weightings in accordance with said assigned roles in accordance with said subject-matter of said database.

119. The method of claim 118, comprising using said importance weightings to discriminate between partially satisfied queries.

120. The method of claim 106, wherein said analysis comprises noun phrase type parsing.

121. The method of claim 106, wherein said analysis comprises using linguistic techniques supported by a knowledge base related to the subject-matter of said stored data items.

122. The method of claim 106, wherein said analysis comprises using statistical classification techniques.

123. The method of claim 106, wherein said analyzing comprises using a combination of:

i) a linguistic technique supported by a knowledge base related to the subject-matter of said stored data items, and

ii) a statistical technique.

124. The method of claim 123, wherein said statistical technique is carried out on a data item following said linguistic technique.

125. The method of claim 123, wherein said linguistic technique comprises at least one of:

segmentation,

tokenization,

lemmatization,

tagging,

part of speech tagging, and

at least partial named entity recognition of said data item.

126. The method of claim 123, further comprising using at least one of probabilities, and probabilities arranged into weightings, to discriminate between different results from said respective techniques.

127. The method of claim 126, further comprising modifying said weightings according to user search behavior.

128. The method of claim 127, wherein said user search behavior comprises past behavior of a current user.

129. The method of claim 127, wherein said user search behavior comprises past behavior aggregated over a group of users.

130. The method of claim 123, wherein an output of said linguistic technique is used as an input to said at least one statistical technique.

131. The method of claim 123, wherein said at least one statistical technique is used within said linguistic technique.

132. The method of claim 123, comprising using two statistical techniques.

133. The method of claim 89, further comprising assigning of at least one code indicative of a meaning associated with at least one of said stored data items, said assignment being to terms likely to be found in queries intended for said at least one stored data item.

134. The method of claim 133, wherein said meaning associated with at least one of said stored data items is at least one of said item, an attribute class of said item and an attribute value of said item.

135. The method of claim 133, further comprising expanding a range of said terms likely to be found in queries by assigning a new term to said at least one code.

136. The method of claim 133, comprising providing groupings of class terms and groupings of attribute value terms.

137. The method of claim 106, wherein, if said analysis identifies an ambiguity, then carrying out a stage of testing said query for semantic validity for each meaning within said ambiguity, and for each meaning found to be semantically valid, presenting said user with a prompt to resolve said validity.

138. The method of claim 106, wherein, if said analysis identifies an ambiguity, then carrying out a stage of testing said query for semantic validity to each meaning within said ambiguity, and for each meaning found to be semantically valid then retrieving data items in accordance therewith and discriminating between said meanings based on corresponding data item retrievals.

139. The method of claim 106, wherein, if said analysis identifies an ambiguity, then carrying out a stage of testing said query for semantic validity to each meaning within said ambiguity, and for each meaning found to be semantically valid, using a knowledge base associated with the subject-matter of said stored data items to discriminate between said semantically valid meanings.

140. The method of claim 89, further comprising predefining for each data item a probability matrix to associate said data item with a set of attribute values.

141. The method of claim 140, further comprising using said probabilities to resolve ambiguities in said query.

142. The method of claim 89, further comprising a stage of processing input text comprising a plurality of terms relating to a predetermined set of concepts, to classify said terms in respect of said concepts, the stage comprising

arranging said predetermined set of concepts into a concept hierarchy,

matching said terms to respective concepts, and

applying further concepts hierarchically related to said matched concepts, to said respective terms.

143. The method of claim 142, wherein said concept hierarchy comprises at least one of the following relationships

(a) a hypernym-hyponym relationship,

(b) a part-whole relationship,

(c) an attribute value dimension—

attribute value relation,

(d) an inter-relationship between neighboring conceptual sub-hierarchies.

144. The method of claim 142, wherein said classifying said terms further comprises applying confidence levels to rank said matched concepts according to types of decisions made to match respective concepts.

145. The method of claim 142, further comprising

identifying prepositions within said text,

using relationships of said prepositions to said terms to identify a term as a focal term, and

setting concepts matched to said focal term as focal concepts.

146. The method of claim 142, wherein said arranging said concepts comprises grouping synonymous concepts together.

147. The method of claim 146, wherein said grouping of synonymous concepts comprises grouping of concept terms being morphological variations of each other.

148. The method of claim 142, wherein at least one of said terms has a plurality of meanings, the method comprising a disambiguation stage of discriminating between said plurality of meanings to select a most likely meaning.

149. The method of claim 148, wherein said disambiguation stage comprises comparing at least one of attribute values, attribute dimensions, brand associations and model associations between said input text and respective concepts of said plurality of meanings.

150. The method of claim 149, wherein said comparing comprises determining statistical probabilities.

151. The method of claim 148, wherein said disambiguation stage comprises identifying a first meaning of said plurality of meanings as being hierarchically related to another of said terms in said text, and selecting said first meaning as said most likely meaning.

152. The method of claim 148, comprising retaining at least two of said plurality of meanings.

153. The method of claim 152, further comprising applying probability levels to each of said retained meanings, thereby to determine a most probable meaning.

154. The method of claim 148, further comprising finding alternative spellings for at least one of said terms, and applying each alternative spelling as an alternative meaning.

155. The method of claim 154, further comprising using respective concept relationships to determine a most likely one of said alternative spellings.

156. The method of claim 142, wherein said input text is an item to be added to a database.

157. The method of claim 142, wherein said input text is a query for searching a database.

158. A query method for searching stored data items, the method comprising:

receiving a query comprising at least a first search term from a user,

expanding the query by adding to said query, terms related to said at least first search term,

analyzing said query for ambiguity,

formulating at least one ambiguity-resolving prompt for said user, such that an answer to said prompt resolves said ambiguity,

modifying said query in view of an answer received to said ambiguity resolving prompt,

retrieving data items corresponding to said modified query,

formulating results-restricting prompts for said user,

selecting at least one of said results-restricting prompts to ask said user, and receiving a response thereto

using said received response to exclude ones of said retrieved items, thereby to provide to said user a subset of said retrieved data items as a query result.

159. The method of claim 158, wherein said query comprises a plurality of terms, and said expanding said query further comprises analyzing said terms to determine a grammatical interrelationship between ones of said terms.

160. The method of claim 158, wherein said expanding comprises a three-stage process of separately adding to said query:

a) items which are closely related to said search term,

b) items which are related to said search term to a lesser degree and

c) an alternative interpretation due to any ambiguity inherent in said search term.

161. The method of claim 158, further comprising at least one additional focusing process of repeating stages iii) to vi), thereby to provide refined subsets of said retrieved data items as said query result.

162. The method of claim 158, further comprising ordering said formulated prompts according to an entropy weighting based on probability values and asking ones of said prompt having more extreme entropy weightings.

163. The method of claim 162, further comprising recalculating said probability values and consequently said entropy weightings following receiving of a response to an earlier prompt.

164. The method of claim 162, further comprising using a dynamic answer set for each prompt, said dynamic answer set comprising answers associated with attribute values, said attribute values being true for some received items and false for other received items, thereby to discriminate between said retrieved items.

165. The method of claim 164, further comprising ranking respective answers within said dynamic answer set according to a respective power to discriminate between said retrieved items.

166. The method of claim 162, further comprising modifying said probability values according to user search behavior.

167. The method of claim 166, wherein said user search behavior comprises past behavior of a current user.

168. The method of claim 166, wherein said user search behavior comprises past behavior aggregated over a group of users.

169. The method of claim 166, wherein said modifying comprises using said user search behavior to obtain a priori selection probabilities of respective data items, and modifying said weightings to reflect said probabilities.

170. The method of claim 162, wherein said entropy weighting is associated with at least one of a group comprising said items, classifications and classification values of respective attributes.

171. The method of claim 158, comprising semantically parsing said stored data items prior to said receiving a query.

172. The method of claim 171, wherein said semantic analysis prior to querying comprises pre-arranging said data items into classes, each class having assigned attribute values, the pre-arranging comprising analyzing said data item to identify therefrom a data item class and if present, attribute values of said class.

173. The method of claim 172, comprising arranging said attribute values into classes.

174. The method of claim 172, wherein said classes are pre-selected for intrinsic meaning to subject matter of a respective database.

175. The method of claim 174, wherein major ones of said classes are arranged hierarchically.

176. The method of claim 173, wherein said attribute classes are arranged hierarchically.

177. The method of claim 176, further comprising determining semantic meaning to a term in said data item from a hierarchical arrangement of said term.

178. The method of claim 175, wherein said classes are also used in analysing said query.

179. The method of claim 174, wherein attribute values are assigned weightings according to the subject-matter of a respective database.

180. The method of claim 174, wherein at least one of said attribute values and said classes are assigned roles in accordance with the subject matter of a respective database.

181. The method of claim 180, wherein said roles are additionally used in parsing said query.

182. The method of claim 181, further comprising assigning importance weightings in accordance with said assigned roles in accordance with said subject-matter.

183. The method of claim 182, comprising using said importance weightings to discriminate between partially satisfied queries.

184. The method of claim 172, wherein said analyzing comprises noun phrase type parsing.

185. The method of claim 172, wherein said analyzing comprises using linguistic techniques supported by a knowledge base related to the subject-matter of said stored data items.

186. The method of claim 172, wherein said analyzing comprises statistical classification techniques.

187. The method of claim 172, wherein said analyzing comprises using a combination of:

i) a linguistic technique supported by a knowledge base related to the subject-matter of said stored data items, and

ii) a statistical technique.

188. The method of claim 187, wherein said statistical technique is carried out on a data item following said linguistic technique.

189. The method of claim 187, wherein said linguistic technique comprises at least one of:

segmentation,

tokenization,

lemmatization,

tagging,

part of speech tagging, and

at least partial named entity recognition of said data item.

190. The method of claim 187, further comprising using at least one of probabilities, and probabilities arranged into weightings, to discriminate between different results from said respective techniques.

191. The method of claim 190, further comprising modifying said weightings according to user search behavior.

192. The method of claim 191, wherein said user search behavior comprises past behavior of a current user.

193. The method of claim 191, wherein said user search behavior comprises past behavior aggregated over a group of users.

194. The method of claim 187, wherein an output of said linguistic technique is used as an input to said at least one statistical technique.

195. The method of claim 187, wherein said at least one statistical technique is used within said linguistic technique.

196. The method of claim 187, comprising using two statistical techniques.

197. The method of claim 158, further comprising assigning of at least one code indicative of a meaning associated with at least one of said stored data items, said assignment being to terms likely to be found in queries intended for said at least one stored data item.

198. The method of claim 197, wherein said meaning associated with at least one of said stored data items is at least one of said item, a classification of said item and classification value of said item.

199. The method of claim 197, further comprising expanding a range of said terms likely to be found in queries by assigning a new term to said at least one code.

200. The method of claim 197, comprising providing groupings of class terms and groupings of attribute value terms.

201. The method of claim 172, wherein, if said analyzing identifies an ambiguity, then carrying out a stage of testing said query for semantic validity for each meaning within said ambiguity, and for each meaning found to be semantically valid, presenting said user with a prompt to resolve said validity.

202. The method of claim 172, wherein, if said analyzing identifies an ambiguity, then carrying out a stage of testing said query for semantic validity to each meaning within said ambiguity, and for each meaning found to be semantically valid then retrieving data items in accordance therewith and discriminating between said meanings based on corresponding data item retrievals.

203. The method of claim 172, wherein, if said analyzing identifies an ambiguity, then carrying out a stage of testing said query for semantic validity to each meaning within said ambiguity, and for each meaning found to be semantically valid, using a knowledge base associated with the subject-matter of said stored data items to discriminate between said semantically valid meanings.

204. The method of claim 158, further comprising predefining for each data item a probability matrix to associate said data item with a set of attribute values.

205. The method of claim 204, further comprising using said probabilities to resolve ambiguities in said query.

206. A query method for searching stored data items, the method comprising:

receiving a query comprising at least two search terms from a user,

analyzing the query by determining a semantic relationship between the search terms thereby to distinguish between terms defining an item and terms defining an attribute value thereof,

retrieving data items corresponding to at least one of identified items,

using attribute values applied to said retrieved data items to formulate prompts for said user,

asking said user at least one of said formulated prompts and receiving a response thereto

using said received response to compare to values of said attributes to exclude ones of said retrieved items, thereby to provide to said user a subset of said retrieved data items as a query result.

207. The method of claim 206, wherein said analyzing the query comprises applying confidence levels to rank said terms according to types of decisions made to reach said terms.

208. A query method for searching stored data items, the method comprising:

receiving a query comprising at least a first search term from a user,

parsing said query to detect noun phrases,

retrieving data items corresponding to said parsed query,

formulating results-restricting prompts for said user,

selecting at least one of said results-restricting prompts to ask a user, and receiving a response thereto

using said received response to exclude ones of said retrieved items, thereby to provide to said user a subset of said retrieved data items as a query result.

209. The query method of claim 208, wherein said parsing comprises identifying:

i) references to stored data items in said query, and

ii) references to at least one of attribute classes and attribute values associated therewith.

210. The query method of claim 209, further comprising assigning importance weights to respective attribute values, said importance weights being usable to gauge a level of correspondence with data items in said retrieving.

211. The query method of claim 208, further comprising ranking said results-restricting prompts and only asking said user highest ranked ones of said prompts.

212. The query method of claim 211, wherein said ranking is in accordance with an ability of a respective prompt to modify a total of said retrieved items.

213. The query method of claim 211, wherein said ranking is in accordance with weightings applied to attribute values to which respective prompts relate.

214. The query method of claim 211, wherein said ranking is in accordance with experience gathered in earlier operations of said method.

215. The query method of claim 214, wherein said experience is at least one of a group comprising experience over all users, experience over a group of selected users, experience from a grouping of similar queries, and experience gathered from a current user.

216. The query method of claim 211, wherein said formulating comprises framing a prompt in accordance with a level of effectiveness in modifying a total of said retrieved items.

217. The query method of claim 211, wherein said formulating comprises weighting attribute values associated with data items of said query and framing a prompt to relate to highest ones of said weighted attribute values.

218. The query method of claim 211, wherein said formulating comprises framing prompts in accordance with experience gathered in earlier operations of said method.

219. The query method of claim 218, wherein said experience is at least one of a group comprising experience over all users, experience gathered from a predetermined group of users, experience gathered from a group of similar queries and experience gathered from a current user.

220. The query method of claim 211, wherein said formulating comprises including a set of at least two answers based on said retrieved results, each answer mapping to at least one retrieved result.

221. An automatic method of classifying stored data relating to a set of objects for a data retrieval system, the method comprising:

defining at least two object classes,

assigning to each class at least one attribute value,

for each attribute value assigned to each class assigning an importance weighting,

assigning objects in said set to at least one class, and

assigning to said object, an attribute value for at least one attribute of said class.

222. The method of claim 221, wherein said objects are represented by textual data and wherein said assigning of objects and assigning of said attribute values comprise using a linguistic algorithm and a knowledge base.

223. The method of claim 221, wherein said objects are represented by textual data and wherein said assigning of objects and assigning of said attribute values comprise using a combination of a linguistic algorithm, a knowledge base and a statistical algorithm.

224. The method of claim 221, wherein said objects are represented by textual data and wherein said assigning of objects and assigning of said attribute values comprise using supervised clustering techniques.

225. The method of claim 224, wherein said supervised clustering comprises initially assigning using a linguistic algorithm and a knowledge base and subsequently adding statistical techniques.

226. The method of claim 221, further comprising providing an object taxonomy within at least one class.

227. The method of claim 221, further comprising providing an attribute value taxonomy within at least one attribute.

228. The method of claim 221, comprising grouping query terms having a similar meaning in respect of said object classes under a single label.

229. The method of claim 221, further comprising grouping attribute values to form a taxonomy.

230. The method of claim 229, wherein said taxonomy is global to a plurality of object classes.

231. The method of claim 221, wherein said objects are represented by textual descriptions comprising a plurality of terms relating to a predetermined set of concepts, the method comprising a stage of analyzing said textual descriptions, to classify said terms in respect of said concepts, the stage comprising

arranging said predetermined set of concepts into a concept hierarchy,

matching said terms to respective concepts, and

applying further concepts hierarchically related to said matched concepts, to said respective terms.

232. The method of claim 231, wherein said concept hierarchy comprises at least one of the following relationships

(a) a hypernym-hyponym relationship,

(b) a part-whole relationship,

(c) an attribute dimension—

attribute value relation,

(d) an inter-relationship between neighboring conceptual sub-hierarchies.

233. The method of claim 231, wherein said classifying said terms further comprises applying confidence levels to rank said matched concepts according to types of decisions made to match respective concepts.

234. The method of claim 231, further comprising

identifying prepositions,

using relationships of said prepositions to said terms to identify a term as a focal term, and

setting concepts matched to said focal term as focal concepts.

235. The method of claim 231, wherein said arranging said concepts comprises grouping synonymous concepts together.

236. The method of claim 235, wherein said grouping of synonymous concepts comprises grouping of concept terms being morphological variations of each other.

237. The method of claim 231, wherein at least one of said terms has a plurality of meanings, the method comprising a disambiguation stage of discriminating between said plurality of meanings to select a most likely meaning.

238. The method of claim 237, wherein said disambiguation stage comprises comparing at least one of attribute values, attribute dimensions, brand associations and model associations between said terms and respective concepts of said plurality of meanings.

239. The method of claim 238, wherein said comparing comprises determining statistical probabilities.

240. The method of claim 237, wherein said disambiguation stage comprises identifying a first meaning of said plurality of meanings as being hierarchically related to another of said terms, and selecting said first meaning as said most likely meaning.

241. The method of claim 237, comprising retaining at least two of said plurality of meanings.

242. The method of claim 241, further comprising applying probability levels to each of said retained meanings, thereby to determine a most probable meaning.

243. The method of claim 237, further comprising finding alternative spellings for at least one of said terms, and applying each alternative spelling as an alternative meaning.

244. The method of claim 243, further comprising using respective concept relationships to determine a most likely one of said alternative spellings.

245. A method of processing input text comprising a plurality of terms relating to a predetermined set of concepts, to classify said terms in respect of said concepts, the method comprising

arranging said predetermined set of concepts into a concept hierarchy,

matching said terms to respective concepts, and

applying further concepts hierarchically related to said matched concepts, to said respective terms.

246. The method of claim 245, wherein said concept hierarchy comprises at least one of the following relationships

(a) a hypernym-hyponym relationship,

(b) a part-whole relationship,

(c) an attribute dimension—

attribute value relation,

(d) an inter-relationship between neighboring conceptual sub-hierarchies.

247. The method of claim 245, wherein said classifying said terms further comprises applying confidence levels to rank said matched concepts according to types of decisions made to match respective concepts.

248. The method of claim 245, further comprising

identifying prepositions within said text,

using relationships of said prepositions to said terms to identify a term as a focal term, and

setting concepts matched to said focal term as focal concepts.

249. The method of claim 245, wherein said arranging said concepts comprises grouping synonymous concepts together.

250. The method of claim 249, wherein said grouping of synonymous concepts comprises grouping of concept terms being morphological variations of each other.

251. The method of claim 245, wherein at least one of said terms comprises a plurality of meanings, the method comprising a disambiguation stage of discriminating between said plurality of meanings to select a most likely meaning.

252. The method of claim 251, wherein said disambiguation stage comprises comparing at least one of attribute values, attribute dimensions, brand associations and model associations between said input text and respective concepts of said plurality of meanings.

253. The method of claim 252, wherein said comparing comprises determining statistical probabilities.

254. The method of claim 251, wherein said disambiguation stage comprises identifying a first meaning of said plurality of meanings as being hierarchically related to another of said terms in said text, and selecting said first meaning as said most likely meaning.

255. The method of claim 251, comprising retaining at least two of said plurality of meanings.

256. The method of claim 255, further comprising applying probability levels to each of said retained meanings, thereby to determine a most probable meaning.

257. The method of claim 251, further comprising finding alternative spellings for at least one of said terms, and applying each alternative spelling as an alternative meaning.

258. The method of claim 257, further comprising using respective concept relationships to determine a most likely one of said alternative spellings.

259. The method of claim 245, wherein said input text is an item to be added to a database.

260. The method of claim 245, wherein said input text is a query for searching a database.