Abstract: The information processing device 1 processes document collections having tags permitting semantic class identification appended to each document and comprises a search unit 2, which creates multiple semantic class units containing one, two, or more semantic classes based on a taxonomy that identifies relationships between semantic classes, and a frequency calculation unit 3 which, for each of the semantic class units, identifies documents that match that semantic class unit in the document collections and, for these matching documents, calculates a first frequency that represents the frequency of occurrence in a designated document collection and a second frequency that represents the frequency of occurrence in non-designated document collections. Once the calculations have been performed, the search unit 2 identifies any of the semantic class units based on the first frequency and the second frequency of the matching documents.

Abstract: In the provided document clustering system (100), a concept tree structure accumulation unit (11) stores a concept tree structure that represents a hierarchical relationship among concepts represented by each of a plurality of words. For any two words, a concept similarity computation unit (12) obtains a concept similarity, which is an index indicating how close the concepts represented by the two words are. Using concept similarities for words that appear in two documents in a document set, an inter-document similarity computation unit (13) obtains an inter-document similarity, which indicates how similar the two documents are semantically. A clustering unit (14) uses inter-document similarities to cluster the documents in the document set.

Abstract: Provided is a related document search system which can provide supplementary information showing a related content together with a related document related to a predetermined document.

Abstract: To provide a technique for structuralizing ontology in a prescribed form to a structure to which features of data are reflected. An ontology processing device has a structuralizing device for structuralizing properties of the ontology in the prescribed form generated from a set of instance data containing a combination of a subject, a property, and an object expressed with a character string according to the features of the object, and has a ontology storage device which stores the ontology structuralized by the structuralizing device. With this structure, the properties of the ontology in the prescribed form are corrected or expressed as an ontology structure by reflecting the characteristics of a set of the objects obtained from the data.

Abstract: An information classification device (1) is provided with an union of sets determination unit (10) which performs correct/incorrect determination regarding a content to be classified using a union of sets rule, and an individual determination unit (11) which applies a plurality of individual determination rules to the content to be classified which has been determined as correct, determines whether the content matches the condition, and performs correct/incorrect determination again regarding the content to be classified which has been determined as correct on the basis of the determination result of each individual determination rule. The union of sets determination rule is created using a result of correct/incorrect determination previously performed by two or more people with respect to a plurality of contents which are different from the contents to be classified, and also using feature amounts of respective different contents.

Abstract: A classification hierarchy regeneration system is provided, wherein when a new classification hierarchy is generated by restructuring an existing classification hierarchy, a classification hierarchy in view of hierarchical relationship of classifications and a classification hierarchy integrating classifications of the same meaning can be efficiently generated. The clustering means clusters a data group associated with a hierarchical classification, and generating a classification group, i.e., a group obtained by extracting a classification satisfying a condition defined in advance from classifications corresponding to respective data in a cluster. The cooccurrence degree calculation means calculates a degree of cooccurrence of two classifications selected from the classification group. The classification hierarchy regeneration means regenerates the hierarchy of classification based on the classification group and the degree of cooccurrence.

Abstract: The disclosed apparatus uses a training data generation apparatus 2, which generates training data used for creating characteristic expression extraction rules. The training data generation apparatus 2 includes: a training data candidate clustering unit 21, which clusters a plurality of training data candidates assigned labels indicating annotation classes based on feature values containing respective context information, and a training data generation unit 22 which, by referring to each cluster obtained using the clustering results, obtains the distribution of the labels of the training data candidates within the cluster, identifies training data candidates that meet a preset condition based on the obtained distribution, and generates training data using the identified training data candidates.

Abstract: A data classifier system of the present invention selects a plurality of classifications correlated to data groups so as to output classification axes based on hierarchical classifications and data groups. The data classifier system includes a basic category accumulation means, a classification axis candidate creation means and a priority calculation means. The basic category accumulation means accumulates classifications serving as basic categories used for selecting desired classifications in advance. The classification axis candidate creation means creates classification axis candidates based on combinations of classifications each correlated to at least one data among descendant classifications of each basic category. The priority calculation means calculates priorities with respect to the classification axis candidates created by the classification axis candidate creation means based on hierarchical distances of classifications in the classified hierarchy.

Abstract: According to the present invention, phrases of the same kind can be extracted from a plurality of documents having various formats. A storage device stores a plurality of documents that have various formats. A pattern candidate creating unit receives a list of input words that are selected as samples among phrases that are to be included in a dictionary. The pattern candidate creating unit selects one document, determines forward and backward character strings of input words in the selected document as candidates of patterns, and stores the forward and backward character strings as a pattern candidate. The pattern candidate creating unit executes the above processes for each of the documents. A phrase candidate creating unit extracts phrases interposed between patterns included in the pattern candidate as candidates of phrases to be output, and stores the extracted phrases as a phrase candidate.

Abstract: A data classifier system of the present invention selects a plurality of classifications correlated to data groups so as to output classification axes based on hierarchical classifications and data groups. The data classifier system includes a basic category accumulation means, a classification axis candidate reduction means and a priority calculation means. The basic category accumulation means accumulates classifications serving as basic categories used for desired classifications in advance. The classification axis candidate reduction means selects a plurality of classifications from among classifications descendant from each basic category so as to create classification axis candidates, thus reducing classification axis candidates subjected to calculations based on data quantity of classifications and hierarchical distances of classifications.

Abstract: A cooccurrence dictionary creating system includes: a language analyzing section which subjects a text to a morpheme analysis, a clause specification, and a modification relationship analysis between clauses, a cooccurrence relationship collecting section which collects cooccurrences of nouns in each clause of the text, modification relationships of nouns and declinable words, and modification relationships between declinable words as cooccurrence relationships, a cooccurrence score calculating section which calculates a cooccurrence score of the cooccurrence relationship based on a frequency of the collected cooccurrence relationship, and a cooccurrence dictionary storage section which stores a cooccurrence dictionary in which a correspondence between the calculated cooccurrence score and the cooccurrence relationship is described.

Abstract: A word classification system is provided with an inter-word pattern learning section for learning at least either the context information or the layout information between classification-known words which co-appear and creating an inter-word pattern for determining whether data relating to a word pair which is a combination of words is data relating to a same-classification word pair which is the combination of words in the same classification or data relating to a different-classification word pair which is a combination of words in different classifications on the basis of the relationship between the classification-known words which co-appear in a document.

Abstract: Sets of strings of which the drawing positions are arranged in one direction are extracted from a document as attribute groups. An attribute name score is calculated for each attribute group to determine an extent to which each attribute group is a set of attribute names. Based on the attribute name scores, an attribute name group is selected out of the attribute groups. From among the attribute groups, an attribute group which includes a string which is the same as at least one string of the attribute name group and of which the drawing position is the same as that of the string of the attribute name group is selected. From the string at the same drawing position, an attribute name is extracted. From the other strings of the selected attribute group than those at the same drawing position, an attribute value corresponding to the attribute name is extracted.

Abstract: An object of the present invention is to perform data clustering while preventing the processing speed from decreasing while maintaining accuracy. A block division section 3 divides a block received from a DB access section 2 into sufficiently small blocks. A block storage section 8 stores blocks supplied from the block division section 3 and hierarchical relationship between the blocks. A block integration section 5 integrates blocks and groups in the order from a hierarchically deeper position to a shallower position based on the stored hierarchical relationship.

Abstract: To provide a technique for structuralizing ontology in a prescribed form to a structure to which features of data are reflected. An ontology processing device has a structuralizing device for structuralizing properties of the ontology in the prescribed form generated from a set of instance data containing a combination of a subject, a property, and an object expressed with a character string according to the features of the object, and has a ontology storage device which stores the ontology structuralized by the structuralizing device. With this structure, the properties of the ontology in the prescribed form are corrected or expressed as an ontology structure by reflecting the characteristics of a set of the objects obtained from the data.

Abstract: A field pair as a combination of a definite field and an indefinite field is decided and a correlation value between the definite field and the indefinite field in each of the field pairs is calculated. Among the field pairs in which the correlation value is not smaller than a threshold value, indefinite fields having corresponding definite fields which belong to the same field group are made to be a new field group.

Abstract: Same document group creation means (11) acquires a ratio of common words and characters between documents in order to obtain a predetermined similarity greater than a predetermined threshold value between the documents. According to the ratio, words or characters are selected with a common priority in all the documents to be matched. The documents are correlated to the same document candidate group identified by the selected words or characters and stored in a same group candidate group storage unit (22).

Abstract: An evaluation polarity of reputation information with an unknown evaluation polarity is estimated by utilizing reputation information with a known evaluation polarity. The present polarity estimation system is a polarity estimation system for estimating an evaluation polarity indicating whether reputation information is positive or negative, and includes a reputation information storage part that precedently stores reputation information with a known evaluation polarity; and a polarity estimating means for estimating an evaluation polarity of reputation information with an unknown evaluation polarity on the basis of the reputation information with the known evaluation polarity precedently stored in the reputation information storage part.

Abstract: An object of the present invention is to perform data clustering while preventing the processing speed from decreasing while maintaining accuracy. A block division section 3 divides a block received from a DB access section 2 into sufficiently small blocks. A block storage section 8 stores blocks supplied from the block division section 3 and hierarchical relationship between the blocks. A block integration section 5 integrates blocks and groups in the order from a hierarchically deeper position to a shallower position based on the stored hierarchical relationship.

Abstract: A device is provided that generates the gestures and expressions of a human image on a computer without expending a great amount of labor. The words for the system response to the input of a user and the state of the dialogue are described in a dialogue flow memory unit, a dialogue flow analysis unit analyzes the spoken text of the flow, extracts the key words associated with a movement pattern by referring to a text movement association table, and the movement expression generation unit generates the movements corresponding to the movement pattern. In the generation of the movement, movement patterns determined in advance are selected according to the state of the dialogue written in the dialogue flow, and the movement pattern is determined or modified by the key words.