WO2006073095A1 - 連想辞書作成装置 - Google Patents
連想辞書作成装置 Download PDFInfo
- Publication number
- WO2006073095A1 WO2006073095A1 PCT/JP2005/023919 JP2005023919W WO2006073095A1 WO 2006073095 A1 WO2006073095 A1 WO 2006073095A1 JP 2005023919 W JP2005023919 W JP 2005023919W WO 2006073095 A1 WO2006073095 A1 WO 2006073095A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- word
- information
- preference
- search
- related word
- Prior art date
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/36—Creation of semantic tools, e.g. ontology or thesauri
Definitions
- the present invention relates to an associative dictionary creating apparatus that creates an associative dictionary storing associations between words in information stored in a database.
- an associative dictionary adapted to the user can also be created by increasing the degree of association between the search keyword entered by the user and the related word candidate ability corresponding to the user-selected related word.
- An associative dictionary creation device that can be created is disclosed (for example, see Patent Document 2).
- Patent Document 1 JP-A-10-134075
- Patent Document 2 Japanese Patent Laid-Open No. 2000-348042
- a conventional associative dictionary creation device that automatically changes the degree of association and creates an associative dictionary adapted to the user when the user selects a related word for the search keyword. Until it was acquired, all search keywords that matched the preference had to be entered, and related terms corresponding to the search keywords had to be selected, and the burden on the user was significant until the accuracy was improved.
- the user is stored in the associative dictionary in order to improve the accuracy of the associative dictionary. Selecting a performer name and a performer associated with the performer name must be performed for multiple performers, and the burden on the user is great.
- the present invention has been made in view of the above circumstances, and provides an associative dictionary creation device capable of easily and efficiently creating an associative dictionary that suits user preferences. Objective.
- an associative dictionary creating apparatus is an associative dictionary creating apparatus that creates an association dictionary that stores relationships between a plurality of words, and is arbitrarily selected from information representing user preferences.
- a word selection means for selecting a word other than the related word search word included in the information representing the user's preference as a related word candidate, and storing information to be searched
- a related word storage that selects a related word related to the related word search word from the related word candidates based on the database, and associates the related word search word with the related word and stores the related word in the associative dictionary Means.
- FIG. 1 is a block diagram showing a configuration of an information search apparatus including an associative dictionary creation apparatus according to Embodiment 1 of the present invention.
- FIG. 2 is a diagram illustrating an example of an associative dictionary stored in an associative dictionary storage unit.
- FIG. 3 is a diagram showing an example of a profile stored in a profile storage unit.
- FIG. 4 is a flowchart showing a flow of operations when creating an associative dictionary in the information retrieval apparatus of the first embodiment.
- FIG. 5 is a flowchart showing the flow of operations when searching for information in the information search apparatus of the first embodiment.
- FIG. 6 is a diagram illustrating an example of an associative dictionary stored in an associative dictionary storage unit.
- FIG. 7 is a block diagram showing a configuration of an information search device including an associative dictionary creation device according to Embodiment 2 of the present invention.
- FIG. 8 is a flowchart showing a flow of operations when creating an associative dictionary in the information search device of the second embodiment.
- FIG. 9 is a diagram showing an example of an associative dictionary stored in an associative dictionary storage unit.
- FIG. 10 is a diagram showing an example of content contents searched by the information search devices of Embodiments 2 and 4.
- FIG. 11 is a diagram showing (a) an example of contents displayed on the display unit and (b) an example of contents stored in the display content storage unit of the information search devices of Embodiments 2 and 4. It is.
- FIG. 12 is a diagram showing (a) an example of contents displayed on the display unit and (b) an example of contents stored in the display content storage unit of the information search devices of Embodiments 2 and 4. It is.
- FIG. 13 is a block diagram showing a configuration of an information search device including an associative dictionary creation device according to Embodiment 3 of the present invention.
- FIG. 14 is a flowchart showing an operation flow when an associative dictionary is created in the information retrieval apparatus of the third embodiment.
- FIG. 15 is a diagram illustrating an example of an associative dictionary stored in an associative dictionary storage unit.
- FIG. 16 is an information search device including an associative dictionary creation device according to Embodiment 3 of the present invention. It is a block diagram which shows the other structure of a device.
- FIG. 17 is a block diagram showing a configuration of an information search device including an associative dictionary creation device according to Embodiment 4 of the present invention.
- FIG. 18 is a flowchart showing an operation flow when an associative dictionary is created in the information retrieval apparatus of the fourth embodiment.
- An associative dictionary creating device is an associative dictionary creating device that creates an associative dictionary that stores relationships between a plurality of words, and includes any word from information representing user preferences.
- a word selection means for selecting a word other than the related word search word included in the information representing the user's preference as a related word candidate, and a database storing information to be searched Based on the related word candidates
- a related word storage means for selecting a related word related to the related word search word from the related word and storing the related word search word and the related word in association with each other. This makes it possible to easily and efficiently create an associative dictionary that matches the user's preference.
- the word selection means may acquire the information representing the user's preference from a profile storing information representing the user's preference. As a result, it is possible to easily obtain information representing the user's preference.
- the related word storage means calculates the relevance between the related word search word and the related word candidate based on the appearance frequency in the database, and the calculated relevance is equal to or greater than a threshold value.
- Word candidates may be selected as related words related to the related word search word. As a result, it is possible to select related words having a high degree of relevance, and to create an associative dictionary storing related words having a high degree of relevance.
- the associative dictionary creation device further includes display information extraction means for extracting information related to the display viewed by the user, and the word selection means is the display extracted by the display information extraction means.
- Information relating to the user's preference is selected, an arbitrary word is selected from the information relating to the display and a related word search word is selected, and words other than the related word search word included in the information related to the display are related You may select it as a word candidate. Thereby, it is possible to acquire information indicating the user's preference regarding the information viewed by the user.
- the associative dictionary creation device further includes search means for searching for information that matches a search condition from the database, and the display viewed by the user is a result of information search, and the display information extraction
- the means may extract information relating to the result of the information search as information relating to the display. As a result, it is possible to acquire information representing the user's preference regarding the result of information retrieval performed by the user.
- the display information extracting means may extract information displayed on the display screen for a predetermined time or more in the information search result as information related to the display. As a result, for example, information displayed for a short time can be excluded without determining what has been browsed by the user.
- the display viewed by the user is a broadcast program, and the display information extracting means
- the information related to the broadcast program may be extracted as the information related to the display. Thereby, an associative dictionary adapted to the user's viewing act can be created.
- the display information extracting means may extract information relating to the broadcast program displayed on the display screen for a predetermined time or more in the broadcast program as information relating to the display. As a result, for example, a broadcast program displayed for a short time can be excluded without determining that the program has been viewed by the user.
- the associative dictionary creation device further extracts preference suitable information from the database as preference preferred information using the profile storing information representing the user's preference as the preference preferred information.
- the related word storage means refers to the preference preference information extracted from the database, selects a related word related to the related word search word from the related word candidates, and the related word
- the search word and the related word may be associated with each other and stored in the associative dictionary. This makes it possible to create an associative dictionary in which the degree of association of words is adapted to the user.
- the word selection unit uses the preference preference information extracted by the preference preference information extraction unit as information indicating the user's preference, selects an arbitrary word from the preference preference information, and relates to the preference preference information.
- a word search word a word other than the related word search word included in the preference suitability information may be selected as a related word candidate. This makes it possible to create an associative dictionary that stores only words that match the user's preference.
- the word selection means obtains information representing the user's preference from a profile storing information representing the user's preference, and selects an arbitrary word from the information representing the user's preference.
- related word search words words other than the related word search words included in the information indicating the user's preference may be selected as related word candidates.
- the associative dictionary creation device further uses display information extraction means for extracting information related to the display viewed by the user, and a profile storing information representing user preferences from the database.
- Preference preference information extraction means for extracting information suitable for user preference as preference preference information
- the word selection means is configured to extract the display information.
- the information related to the display extracted by the means is set as information representing the user's preference, and the related word search word included in the information related to the display is selected as a related word search word from the information related to the display.
- the related word storage means refers to the preference suitable information extracted from the database and relates to the related word search word from the related word candidates.
- a related word may be selected, and the related word search word and the related word may be associated and stored in the associative dictionary.
- the present invention can be realized not only as an associative dictionary creating apparatus, but also as an associative dictionary creating method using characteristic means included in such an associative dictionary creating apparatus as a step. It can also be realized as a program that causes a computer to execute these steps. Needless to say, such a program can be distributed via a recording medium such as a CD-ROM or a transmission medium such as the Internet.
- FIG. 1 is a block diagram showing a configuration of an information search apparatus including an associative dictionary creation apparatus according to Embodiment 1 of the present invention.
- This information retrieval apparatus is an apparatus for creating an associative dictionary storing associations between words and performing information retrieval using the associative dictionary.
- the content storage unit 101 An associative dictionary storage unit 102, an associative dictionary creation unit 103, an input unit 107, and an associative search unit 108.
- the content storage unit 101 stores information related to television broadcast programs, information related to content stored in a node disc recorder, and the like.
- the associative dictionary storage unit 102 stores an associative dictionary that represents the relationship between words reflecting the user's preference with respect to a plurality of words in the content stored in the content storage unit 101.
- FIG. 2 is a diagram illustrating an example of an associative dictionary stored in the associative dictionary storage unit 102. In the associative dictionary storage unit 102, for example, as shown in FIG.
- the related word pair (related word 1, related word 2) and the related degree indicating the related degree of the related word pair are stored in association with each other, and (related word pair, related degree) is ((Matsushita Taro, Jiro Matsushita), 8 0), ((Taro Matsushita, Hanako Matsushita), 70), ((Jiro Matsushita, Hanako Matsushita), 90), ((Jiro Matsushita, Saburo Matsushita), 80), ... , Is stored.
- the associative dictionary creation unit 103 calculates the relevance of the word pair to be extracted from the content stored in the content storage unit 101, which is information relating to the user's preference, and extracts the word pair to be calculated.
- an association dictionary storage unit 102 that associates the calculated degrees of association with each other and stores them in the association dictionary storage unit 102, and includes a profile storage unit 104, a word selection unit 105, and an association degree calculation unit 106.
- the profile storage unit 104 stores a profile that is information relating to user preferences.
- the profile includes the genre to which the user's favorite content belongs, favorite title and performer, viewing time, etc., and is associated with each data (genre, title, performer, viewing time, etc.).
- a weight value indicating the degree of preference is stored.
- FIG. 3 is a diagram illustrating an example of a profile stored in the profile storage unit 104. In the profile, for example, as shown in FIG. 3, a label indicating the type of data, a data and a weight value indicating the degree of preference of the user with respect to the data are stored in association with each other (label, data, weight value).
- the word selection unit 105 selects words from the profile file stored in the profile storage unit 104 at predetermined time intervals as related word search words, and further includes words other than the related word search words included in the profile. Is selected as a related word candidate.
- the word selection unit 10 5 selects data from the profile of FIG. 3 stored in the profile storage unit 104 as a related word search word, and further selects data other than the related word search word included in the profile as a related word candidate. Select as. Further, only related data having a weight value equal to or greater than a specific threshold may be used as related word search words and related word candidates.
- Relevance calculation unit 106 refers to the content stored in content storage unit 101. Then, the degree of association between the related word search word selected by the word selection unit 105 and the word of the related word candidate is calculated, and when the degree of association is greater than a predetermined threshold, the word of the related word search word and the related word candidate Are stored in the associative dictionary storage unit 102 in association with the related word pairs and the degree of association.
- a method for calculating the degree of association for example, there is a method using mutual information (Ml).
- a is the number of content that includes word W1 and word W2
- b is the number of content that does not include word W1 and includes word W2
- c is the number of content that includes word W1 and does not include word W2
- MI log2 (aN / ((a + b) (a + c) It can be obtained by)).
- Dice—coefficien t weighted Dice—coefficient t-score,% square value, log likelihood ratio, and the like.
- the relevance calculation unit 106 refers to the content stored in the content storage unit 101, and performs word search for each word of the related word search word and the related word candidate selected by the word selection unit 105.
- the text before and after is divided into morphemes and words, and the presence of morphemes and words, the number of appearances, the appearance frequency (tf value) multiplied by the idf (inverse document frequency) value, etc. are used as components.
- a vector may be generated, and a cosine distance between the vector of related word search words and each word vector of related word candidates, the reciprocal of the Euclidean distance, or the like may be calculated as the degree of association.
- the input unit 107 includes an input device such as a keyboard, a mouse, and a remote controller, and the user inputs search conditions and input data including a search keyword.
- the associative search unit 108 acquires the search condition input from the input unit 101 and related terms related to the search keyword included in the search condition from the associative dictionary storage unit 102, generates a search expansion condition, and searches A processing unit that acquires content that matches the extended condition from the content storage unit 101 and outputs it as a search result, and includes a search condition generation unit 109 and a search unit 110.
- the search condition generation unit 109 acquires and acquires a related word that is associated with a search keyword included in the search condition input from the input unit 101 with a degree of relevance greater than or equal to a predetermined threshold.
- a search expansion condition is generated from the related words and the search keywords.
- the search unit 110 is content that matches the search condition generated by the search condition generation unit 109. Is obtained from the content storage unit 101 and used as a search result.
- FIG. 4 is a flowchart showing the flow of operations when creating an associative dictionary.
- the word selection unit 105 selects words from the profile file stored in the profile storage unit 104 at predetermined time intervals to obtain related word search words, and further includes words other than the related word search words included in the profile. Is selected as a related word candidate (step S101). As a specific example, the word selection unit 105 searches related words for the words “Taro Matsushita, Jiro Matsushita,...” From the profile of FIG. 3 stored in the profile storage unit 104 every week as a predetermined time.
- the related degree calculating unit 106 refers to the content stored in the content storage unit 101, and calculates the related degree between the related word search word selected by the word selecting unit 105 and the related word candidate word. (Step S102), it is determined whether or not the relevance is greater than a predetermined threshold (Step S103). As a result, if the degree of relevance is greater than the predetermined threshold (YES in step S103), the related word search word and the word of the related word candidate are used as a related word pair, and the related word pair is associated with the related degree. And stored in the associative dictionary storage unit 102 (step S104). On the other hand, if the relevance is less than or equal to the predetermined threshold (NO in step S103), no processing is performed.
- Step S106 the relevance calculation unit 106 refers to the content stored in the content storage unit 101 and searches each related word search word “Taro Matsushita, Jiro Matsushita, "! for each related word candidate "Jiro Matsushita, Saburo Matsushita, Hanako Matsushita, Drama, 12:00 to 13:00, ", "Taro Matsushita, Saburo Matsushita, Hanako Matsushita, Drama , 12:00 to 13:00, ⁇ ", ..., the degree of association with each word is calculated.
- (related word search word, related word candidate word, relevance) is (Matsushita Taro, Jiro Matsushita, 80), (Taro Matsushita, Saburo Matsushita, 40), (Taro Matsushita, Hanako Matsushita, 70), (Taro Matsushita, Drama, 20), (Taro Matsushita, 12:00 to 13:00, 30 ), (Jiro Matsushita, Taro Matsushita, 80), (Jiro Matsushita, Saburo Matsushita, 80), (Jiro Matsushita, Hanako Matsushita, 90), (Jiro Matsushita, Drama, 10), (Jiro Matsushita, 12: 00 ⁇ 13:00, 40), ..., and the association degree is greater than a predetermined threshold value (for example, 50 is assumed) (related word search word, related word candidate word, association degree) Store
- FIG. 5 is a flowchart showing the flow of operations when searching for information.
- Search condition generation unit 109 receives the search condition input from input unit 107 by the user (step S201).
- the search condition “Taro Matsushita's program” is input from the input unit 107 by the user as an example.
- the search condition generation unit 109 sends from the associative dictionary storage unit 102 related words that are associated with the search keyword included in the search condition input from the input unit 101 with a degree of relevance equal to or greater than a predetermined threshold. Obtain (Step S202). Then, the search condition generation unit 109 generates a search expansion condition from the acquired related words and the search keyword (step S203). In the above example, the search condition generation unit 109 uses the search keyword “Taro Matsushita” included in the search condition “Taro Matsushita's program” input from the input unit 101 and a predetermined threshold (for example, the predetermined threshold is 75).
- the related word “Jiro Matsushita” associated with the above relevance is acquired from the associative dictionary storage unit 102, and either of the acquired related word “Jiro Matsushita” or the search keyword “Taro Matsushita” A search expansion condition “Taro Matsushita + Jiro Matsushita” is generated to indicate that it is included.
- the search unit 110 acquires content that matches the search condition generated by the search condition generation unit 109 from the content storage unit 101 and uses it as a search result (step S 204).
- the search unit 110 uses the search condition “Taro Matsushita + Jiro Matsushita” generated by the search condition generation unit 109 to include a keyword including either “Taro Matsushita” or “Jiro Matsushita”.
- the content is acquired from the content storage unit 101 and used as a search result.
- the word selection unit 105 selects the related word search word and the related word candidate from the profile stored in the profile storage unit 104 every predetermined time.
- the related word search word and the related word candidate may be selected every time the profile is updated instead of every time. By doing this, it is possible to acquire related word pairs without causing a time difference in the change of user's preference.
- the relevance calculation unit 106 calculates the relevance for the related word search word selected by the word selection unit 105 and all words of the related word candidates, and the relevance is If it is greater than the predetermined threshold value, the related word search word and the related word candidate word are used as a related word pair, and the related word pair and the related degree are stored in association in the associative dictionary storage unit 102. It is not something you can. For example, when calculating the related word pair newly selected by the word selecting unit 105 and the related degree, the related degree calculating unit 106 attenuates all the related degrees stored in the associative dictionary storage unit 102.
- association degree and the degree of association calculation unit 106 are already stored in the association dictionary storage unit 102. It may be updated by adding the newly obtained relevance level.
- Relevance Calculation Unit 106 (Taro Matsushita, Jiro Matsushita, 60)
- the relevance stored in the part 102 is attenuated by multiplying the attenuation parameter (for example, 0.8), and the relevance 64 after attenuation of the existing related word pair (Taro Matsushita, Jiro Matsushita)
- a value 124 obtained by adding the newly obtained relevance 60 is stored in the associative dictionary storage unit 102 as the relevance of the related word pair (Taro Matsushita, Jiro Matsushita).
- the contents of the associative dictionary storage unit 102 are updated as shown in FIG. By doing so, it is possible to obtain a degree of relevance that takes into account the degree of relevance with respect to past content, and to improve the accuracy of the relevance.
- the degree of association is calculated for the related word search word selected by the word selection unit 105 and all words of the related word candidate, and whether or not the degree of association is greater than a predetermined threshold value.
- the related word search word and the related word candidate word may be used as a related word pair, and the related word pair and the related degree may be associated and stored in the associative dictionary storage unit 102. Note that, as in this embodiment, when the relevance is greater than a predetermined threshold, the associated word pair and the relevance are associated with each other and the associative dictionary is recorded. If stored in the storage unit 102, the storage capacity can be reduced.
- the profile association dictionary which is information representing the user's preference
- an association that stores only the word that matches the user's preference You can create a dictionary.
- the search conditions can be expanded by using words that match the user's preference by using the created associative dictionary. Therefore, the expanded search result can be adapted to the user's preference.
- FIG. 7 is a block diagram showing a configuration of an information search device including the associative dictionary creation device according to Embodiment 2 of the present invention.
- components having the same reference numerals as those in Embodiment 1 perform the same operation, and detailed description thereof will be omitted.
- the information search apparatus of the present embodiment is an apparatus for creating an associative dictionary storing associations between words and performing an information search using the associative dictionary, as shown in FIG.
- the content storage unit 101, the association dictionary storage unit 102, the input unit 107, the association search unit 108, the association dictionary creation unit 201, the display unit 204, the screen creation unit 205, and the broadcast wave reception unit 206 are configured.
- the display unit 204 includes a CRT display, a liquid crystal display (LCD), and a plasma display.
- the broadcast wave receiving unit 206 receives content such as a television broadcast program via the antenna 207.
- Screen creation unit 205 converts the search result searched by search unit 110 into display screen information, outputs the display screen information to display unit 204, and receives the television broadcast received by broadcast wave reception unit 206. Content such as a program is output to the display unit 204. Furthermore, the screen creation unit 205 extracts text information related to the search result searched by the search unit 110 and information related to content such as a television broadcast program displayed on the display unit 204 as information related to display, and the display content storage unit 202 To store.
- the associative dictionary creation unit 201 calculates the relevance of the word pairs extracted from the displayed search results for the content stored in the content storage unit 101, and the word pairs to be calculated are calculated.
- An association dictionary that stores the association degree calculated in association with the association dictionary storage unit 102
- the device includes a display content storage unit 202, a word selection unit 203, and an association degree calculation unit 106.
- the display content storage unit 202 stores text information related to the displayed search results and information related to display that is information related to the displayed content. Each time new text information is stored in the display content storage unit 202, the word selection unit 203 extracts a word by performing morphological analysis on the text information, and extracts a word from the extracted word. A plurality of sets are generated by selecting the related word search words and selecting the extracted words other than the related word search words as related word candidates.
- the degree-of-association calculation unit 106 refers to the content stored in the content storage unit 101 and calculates the degree of association between the related word search word selected by the word selection unit 203 and the word of the related word candidate.
- the degree of relevance is greater than a predetermined threshold
- the related word search word and the word of the related word candidate are used as a related word pair, and the related word pair and the related degree are associated and stored in the associative dictionary storage unit 102.
- a method for calculating the degree of association the same method as in the first embodiment is used.
- Fig. 8 is a flowchart showing the flow of operations when creating an associative dictionary.
- the screen creation unit 205 converts the search result searched by the search unit 110 into display screen information, outputs the display screen information to the display unit 204, or receives the TV received by the broadcast wave receiving unit 206. It is assumed that content such as broadcast programs is output to the display unit 204.
- the search condition “Taro Matsushita's program” is entered by the user, and the video content in which the search result power ⁇ Taro Matsushita appears is explained.
- content such as a TV broadcast program it is assumed that content having “appearance: Jiro Matsushita, Hanayo Matsushita” is displayed as text information related to the content.
- the screen creation unit 205 displays text information related to the search result and the content received by the broadcast wave reception unit 206 when the search result searched by the search unit 110 is displayed.
- the text information related to the content is stored in the display content storage unit 202 as information related to display (step S301).
- the broadcast wave receiving unit 206 When the received content is displayed, text information related to the content may be stored in the display content storage unit 202 as long as the content is displayed for a predetermined time (for example, 10 minutes).
- the screen creation unit 205 displays the text “Appearance: Matsushita” that describes “Taro Matsushita” searched by the search unit 110 and the video content.
- “Taro, Hanako Matsushita, Saburo Matsushita” are stored in the display content storage unit 202. Further, when the content received by the broadcast wave receiving unit 206 in the above example is displayed, “appearance: Jiro Matsushita, Hanayo Matsushita” is stored in the display content storage unit 202.
- the word selection unit 203 extracts a word by performing morphological analysis on the text information and extracts the extracted word. From the extracted words, a word is selected as a related word search word, and a plurality of sets are generated using the extracted words other than the related word search word as related word candidates (step S302). For example, if the text “Appearance: Taro Matsushita, Hanako Matsushita, Saburo Matsushita” is stored in the display content storage unit 202, the word selection unit 203 will respond to “Appearance: Taro Matsushita, Hanako Matsushita, Saburo Matsushita”.
- the words “Taro Matsushita, Hanako Matsushita, Saburo Matsushita” that represent names are extracted as (Related Word Search Word, (Related Word Candidate)) (Taro Matsushita, (Hanako Matsushita, Matsushita) Saburo), (Hanako Matsushita, (Taro Matsushita, Saburo Matsushita)), (Saburo Matsushita, (Taro Matsushita, Hanako Matsushita)).
- the degree-of-association calculation unit 106 calculates the degree of association between the related word search word selected by the word selection unit 105 and the word of the related word candidate with reference to the content stored in the content storage unit 101. (Step S303), it is determined whether or not the relevance is greater than a predetermined threshold (Step S304). As a result, when the relevance is greater than the predetermined threshold (YES in step S304), the related word search word and the related word candidate word are used as related word pairs, and the related word pairs are associated with the related degrees. They are stored in the associative dictionary storage unit 102 (step S305).
- related word pairs for which new association degrees have been calculated may overlap because one word may be a related word search word or a related word candidate word.
- the relevancy calculation unit 106 refers to the content stored in the content storage unit 101 and the string selected by the word selection unit 105 (related word search word, (related word candidate)).
- (related word search word, related word candidate word, relevance) are (Taro Matsushita, Hanako Matsushita, 60), (Taro Matsushita, Saburo Matsushita, 40), (Hanako Matsushita, Taro Matsushita, 60), (Hanako Matsushita, Saburo Matsushita, 70), (Saburo Matsushita, Taro Matsushita, 40), (Saburo Matsushita, Hanako Matsushita, 70) and the relevance is greater than a predetermined threshold (50 as an example)
- a predetermined threshold 50 as an example
- the related word search word, related word candidate word, and related degree are attenuated by multiplying the related degree stored in the association dictionary storage unit 102 shown in FIG. 2 by an attenuation parameter (for example, 0.8).
- an attenuation parameter for example, 0.8
- the association degree calculation unit 106 when updating the association dictionary storage unit 102, the association degree calculation unit 106 attenuates all association degrees stored in the association dictionary storage unit 102, and then newly If a related word pair that has the same relationship as the related word pair already exists in the associative dictionary storage unit 102, a new one is obtained by the relevance and relevance calculation unit 106 already stored in the associative dictionary storage unit 102.
- the power that is updated by adding up the degree of relevance obtained is not limited to this.
- the degree-of-association calculation unit 106 may store the related word pairs for which the degree of association is newly calculated in the associative dictionary storage unit 102 in association with the degree of association.
- the moving image content is searched.
- this is not a limitation, and it may be a hypertext or a web page on the Internet.
- the screen creation unit 205 stores the hypertext and web page display text searched by the search unit 110 in the display content storage unit 202, and further displays the searched noyper text and web page text. Convert to screen information and output the display screen information to the display unit 204.
- the screen creation unit 205 reads from the file “name.html”.
- the display text “Taro Matsushita, Hanako Matsushita, Saburo Matsushita, Jiro Matsushita” shown in FIG. 11 (b) is stored in the display content storage unit 202, and the file “name.html” is converted into display screen information. Display screen information is output to the display unit 204 as shown in FIG. Furthermore, when all the text is not displayed on the display unit 204, the screen creation unit 205 may store only the displayed text in the display content storage unit 202. In the above embodiment, as an example, when the file “name.htmlj described in the HTML language of FIG. 10 is displayed as shown in FIG.
- the display content storage unit 202 may treat the text displayed on the display unit 204 for a predetermined time or more as the displayed text. This eliminates text that is only displayed when scrolling and is not viewed by the user.
- the screen creation unit 205 stores text information related to the content searched by the search unit 110 in the display content storage unit 202, but the content is a video content and corresponds to each scene or frame of the video content.
- the screen creation unit 205 may store only the text corresponding to the scene or frame of the moving image content viewed by the user in the display content storage unit 202. In this way, it is possible to create an associative dictionary that only uses text corresponding to the scenes and frames of the video content that the user has watched. Can be created.
- an associative dictionary adapted to a user's viewing act is created by selecting a word to be stored in the associative dictionary from text related to the content viewed by the user. be able to. Furthermore, when the user searches for content, the search condition can be expanded by using a word that matches the viewing behavior of the user by using the created associative dictionary. Therefore, the extended search result can be adapted to the viewing behavior by the user. [0063] (Embodiment 3)
- FIG. 13 is a block diagram showing a configuration of an information search apparatus including the associative dictionary creation apparatus according to Embodiment 3 of the present invention.
- components having the same reference numerals as those in Embodiment 1 perform the same operation, and detailed description thereof will be omitted.
- the information search apparatus of the present embodiment is an apparatus for creating an associative dictionary storing associations between words and performing an information search using the associative dictionary, as shown in FIG.
- the content storage unit 101, the associative dictionary storage unit 102, the input unit 107, the associative search unit 108, and the associative dictionary creation unit 301 are configured.
- the associative dictionary creation unit 301 extracts preference preference information as text relating to content that matches the preference of the content browser stored in the content storage unit 101, and further, for the preference preference information
- An association dictionary creation device that calculates the degree of association between word pairs extracted from preference preference information, associates the calculated word pair with the calculated degree of association, and stores the association in the association dictionary storage unit 102.
- Unit 104, preference content extraction unit 302, preference preference information storage unit 303, word selection unit 304, and relevance calculation unit 305 is a dictionary storage unit 102.
- the preference preference information storage unit 303 stores preference preference information that is text related to content that matches the user's preference.
- the preference content extraction unit 302 is a text relating to the content that matches the user's preference using the profile, which is the information about the user's preference stored in the profile storage unit 104, as well as the content power stored in the content storage unit 101.
- the preference preference information is extracted and stored in the preference preference information storage unit 303.
- the word selection unit 304 extracts a word by performing morphological analysis on the preference suitability information stored in the preference suitability information storage unit 303, selects a word from the extracted word, and selects a related word.
- the search word is generated, and a plurality of sets of extracted words other than the related word search word are used as related word candidates.
- the degree-of-association calculation unit 305 refers to the preference preference information stored in the preference preference information storage unit 303 and relates the relationship between the related word search word selected by the word selection unit 304 and the word of the related word candidate.
- the degree of relevance is greater than a predetermined threshold
- the related word search word and the word of the related word candidate are used as a related word pair, and the related word pair and the degree of association are associated with each other and the associative dictionary storage unit Store in 102.
- a method similar to that in the first embodiment is used as a method for calculating the degree of association.
- FIG. 14 is a flowchart showing the flow of operations when creating an associative dictionary.
- the preference content extraction unit 302 uses the profile shown in FIG. 3, for example, to store the content power stored in the content storage unit 101 and the text relating to the content that matches the user's preference.
- the preference preference information is extracted and stored in the preference preference information storage unit 303 (step S401).
- the preference content extraction unit 302 acquires (label, data, weight value) from the profile storage unit 104, and for each text related to each content stored in the content storage unit 101. If the text contains a keyword for the data, the weight for the keyword is added, and if the added result exceeds a predetermined threshold, the text exceeding the threshold is stored in the preference information storage unit 303. To do.
- the preference content extraction unit 302 extracts the text “Genre: Drama, Appearance: Taro Matsushita, Hanako Matsushita, Saburo Matsushita” from the content storage unit 101 and stores it in the preference preference information storage unit 303 as preference preference information. Pay.
- the word selection unit 304 extracts a word by performing morphological analysis on the preference suitability information stored in the preference suitability information storage unit 303, selects a word from the extracted word, and selects a related word. A plurality of sets of search words and further extracted words other than related word search words are generated as related word candidates (step S402).
- the word selection unit 30 4 performs morphological analysis on the text “Genre: Drama, Appearance: Taro Matsushita, Hanako Matsushita, Saburo Matsushita” stored in the preference preference information storage unit 303, for example,
- the words “Taro Matsushita, Hanako Matsushita, Saburo Matsushita” representing the name are extracted, and (Taro Matsushita, (Hanako Matsushita, Saburo Matsushita)), (Hanako Matsushita, ( Taro Matsushita, Saburo Matsushita)), (Saburo Matsushita, (Taro Matsushita, Hanako Matsushita))).
- the degree-of-association calculation unit 305 refers to the preference preference information stored in the preference preference information storage unit 303 and relates the relationship between the related word search word selected by the word selection unit 304 and the word of the related word candidate.
- the degree of association is calculated (step S403), and it is determined whether or not the degree of association is greater than a predetermined threshold (step S404). As a result, if the degree of relevance is greater than the predetermined threshold (YES in step S404), the related word search word and the word of the related word candidate are used as a related word pair, and the related word pair is associated with the related degree. It stores in the associative dictionary storage unit 102 (step S405).
- the related word pairs for which the degree of relevance is newly calculated may be duplicated because one word may be a related word search word and another related word candidate word. And stored in the associative dictionary storage unit 102. On the other hand, if the relevance is less than or equal to a predetermined threshold (NO in step S404), no processing is performed. These processes (steps S403 to S405) are performed for all words included in the related word candidates (step S406), and this process (steps S403 to S406) is performed for all related word search words. (Step S40 7).
- the relevance calculation unit 305 refers to the preference preference information stored in the preference preference information storage unit 303 and is selected by the word selection unit 304 (related word search word, (related word candidate) ) (Taro Matsushita, (Hanako Matsushita, Saburo Matsushita)), (Hanako Matsushita, (Taro Matsushita, Saburo Matsushita)), (Saburo Matsushita, (Taro Matsushita, Hanako Matsushita)) Related words The degree of association between the search word and each word of the related word candidate is calculated.
- (related word search word, related word candidate word, relevance) are (Taro Matsushita, Hanako Matsushita, 60), (Taro Matsushita, Saburo Matsushita, 40), (Hanako Matsushita, Taro Matsushita, 60) , (Hanako Matsushita, Saburo Matsushita, 70), (Saburo Matsushita, Taro Matsushita, 40), (Saburo Matsushita, Hanako Matsushita, 70), and the relevance is more than the predetermined threshold (50 as an example)
- a large (related word search word, related word candidate word, degree of association) is stored in the associative dictionary storage unit 102.
- the contents of the associative dictionary storage unit 102 in this case are as shown in FIG.
- the relevance calculation unit 305 calculates the relevance for the related word search word selected by the word selection unit 304 and all the words of the related word candidates. If it is larger than the predetermined threshold, the related word search word and the related word candidate word are stored as related word pairs in the associative dictionary storage unit 102 in association with each other, but the present invention is not limited to this. Is not something For example, the relevance calculation unit 305 is newly selected by the word selection unit 304. When calculating the related word pair and its related degree, after the total related degree stored in the associative dictionary storage unit 102 is attenuated, the same related word pair as the related word pair for which the new related degree is calculated is obtained.
- the degree-of-association calculation unit 305 refers to the preference suitability information stored in the preference preference information storage unit 303 and refers to the preference suitability information that is selected by the word selection unit 304 (related word search word, (related word Candidates)) (Taro Matsushita, (Hanako Matsushita, Saburo Matsushita)), (Hanako Matsushita, (Taro Matsushita, Saburo Matsushita)), (Saburo Matsushita, (Taro Matsushita, Hanako Matsushita))))
- the degree of association between the related word search word and each word of the related word candidate is calculated.
- (related word search word, related word candidate word, relevance) are (Taro Matsushita, Hanako Matsushita, 60), (Taro Matsushita, Saburo Matsushita, 40), (Hanako Matsushita, Taro Matsushita, 60) , (Hanako Matsushita, Saburo Matsushita, 70), (Saburo Matsushita, Taro Matsushita, 40), (Saburo Matsushita, Hanako Matsushita, 70), and the relevance is more than the predetermined threshold (50 as an example) 2, multiply the relevance stored in the associative dictionary storage unit 102 shown in FIG.
- an attenuation parameter for example, 0.8. It is added to the associative dictionary storage unit 102 after being attenuated.
- the contents of the associative dictionary storage unit 102 in this case are as shown in FIG. In this way, it is possible to obtain a degree of relevance that takes into account the degree of relevance to past content, and to improve the accuracy of the relevance.
- the word selection unit 304 selects the related word search word and the related word candidate from the preference suitable information storage unit 303.
- the related word search word and the related word candidate may be selected from the profile storage unit 104.
- the content ability associating with the user's preference The word stored in the associative dictionary and calculating the degree of relevance match the user's preference.
- An associative dictionary that stores only words and adapts the degree of word relevance to the user Can be created.
- the created association dictionary is used to make associations that match the user's preference and to expand search conditions using words that match the user's preference. it can. Therefore, the expanded search result can be further adapted to the user's preference.
- FIG. 17 is a block diagram showing a configuration of an information search device including the associative dictionary creation device according to Embodiment 4 of the present invention.
- components having the same reference numerals as those in the first to third embodiments perform the same operation, and detailed descriptions thereof are omitted.
- the information search apparatus of the present embodiment is an apparatus for creating an associative dictionary storing associations between words and performing an information search using the associative dictionary. As shown in FIG.
- the content storage unit 101, the association dictionary storage unit 102, the input unit 107, the association search unit 108, the display unit 204, the screen creation unit 205, the broadcast wave reception unit 206, and the association dictionary creation unit 401 are configured.
- the associative dictionary creation unit 401 extracts preference preference information as text related to content that matches the preference of the content browser stored in the content storage unit 101, and further, for the preference preference information
- the associative dictionary creation device that calculates the relevance of the word pairs to be extracted from the displayed search results and associates the calculated word pairs with the calculated relevance and stores them in the associative dictionary storage unit 102.
- the preference preference information storage unit 303 stores preference preference information that is text related to content that matches the user's preference.
- the preference content extraction unit 302 is a text relating to the content that matches the user's preference using the profile, which is the information about the user's preference stored in the profile storage unit 104, as well as the content power stored in the content storage unit 101.
- the preference preference information is extracted and stored in the preference preference information storage unit 303.
- the display content storage unit 202 stores text information related to the displayed search results and information related to display that is information related to the displayed content. Each time new text information is stored in the display content storage unit 202, the word selection unit 203 A word is extracted by performing morphological analysis etc. on the information, a word is selected from the extracted word as a related word search word, and a word other than the related word search word is extracted from the extracted word. Generate multiple pairs as candidates.
- the degree-of-association calculation unit 305 refers to the preference preference information stored in the preference preference information storage unit 303 and relates the relationship between the related word search word selected by the word selection unit 304 and the related word candidate word.
- the related degree is greater than a predetermined threshold, the related word search word and the related word candidate word are used as related word pairs, and the related word pair and the related degree are associated with each other, and the associative dictionary storage unit 102 To store.
- a method similar to that in the first embodiment is used as a method for calculating the degree of association.
- FIG. 18 is a flowchart showing the flow of operations when creating an associative dictionary.
- the screen creation unit 205 converts the search result searched by the search unit 110 into display screen information, and outputs the display screen information to the display unit 204 or broadcast wave as in the second embodiment. It is assumed that content such as a TV broadcast program received by the receiving unit 206 is output to the display unit 204.
- the search result is displayed, the search condition “Taro Matsushita's program” is entered by the user, and the search result is a text that describes the video content and video content in which “Taro Matsushita” appears. “Performance: Taro Matsushita, Hanako Matsushita, Saburo Matsushita”.
- a content such as a TV broadcast program it is assumed that the content having “appearance: Jiro Matsushita, Hanayo Matsushita” as text information about the content is displayed.
- the screen creation unit 205 displays the search result searched by the search unit 110 as in the second embodiment, and in this case, the broadcast wave receiving unit 206 displays text information related to the search result.
- the text information related to the content is stored in the display content storage unit 202 as information related to display (step S501). If the content received by the broadcast wave receiving unit 206 is displayed, text information about the content is stored in the display content storage unit 202 if the content is displayed for a predetermined time (for example, 10 minutes). It's okay to do it.
- the screen creation unit 205 displays the text “Appearance: Taro Matsushita, Hanako Matsushita, Saburo Matsushita” which is a description of the video content in which “Taro Matsushita” searched by the search unit 110 appears. Store in the storage unit 202. Further, when the content received by the broadcast wave receiving unit 206 in the above example is displayed, “appearance: Jiro Matsushita, Hanayo Matsushita” is stored in the display content storage unit 202.
- the preference content extraction unit 302 uses the profile as shown in FIG. 3, for example, in which the content power stored in the content storage unit 101 is also stored in the profile storage unit 104 as in the third embodiment. Then, preference preference information, which is text relating to content that matches the user's preference, is extracted and stored in the preference preference information storage unit 303 (step S502). As a specific example, the preference content extraction unit 302 acquires (label, data, weight value) from the profile storage unit 104, and for each text related to each content stored in the content storage unit 101, When a text keyword is included in the text, the weight for the keyword is added. When the result of addition exceeds a predetermined threshold, the text exceeding the threshold is stored in the preference information storage unit 303.
- the preference content extraction unit 302 extracts the text “Genre: Drama, Appearance: Taro Matsushita, Hanako Matsushita, Saburo Matsushita”,... From the content storage unit 101 and stores preference preference information as preference preference information. Stored in part 303.
- the word selection unit 203 extracts a word by performing morphological analysis or the like on the text information, and extracts the extracted word information.
- a word is selected from the selected words and used as a related word search word, and a plurality of sets are generated using the extracted words other than the related word search word as related word candidates (step S503).
- the word selection unit 203 will respond to “Appearance: Taro Matsushita, Hanako Matsushita, Saburo Matsushita”.
- the words “Taro Matsushita, Hanako Matsushita, Saburo Matsushita” that represent names are extracted as (Related Word Search Word, (Related Word Candidate)) (Taro Matsushita, (Hanako Matsushita, Matsushita) Saburo), (Hanako Matsushita, (Taro Matsushita, Saburo Matsushita)), (Saburo Matsushita, (Taro Matsushita, Hanako Matsushita)).
- the degree-of-association calculation unit 305 refers to the preference preference information stored in the preference preference information storage unit 303. Then, the degree of association between the related word search word selected by the word selection unit 203 and the word of the related word candidate is calculated (step S504), and it is determined whether or not the degree of association is greater than a predetermined threshold. (Step S505). As a result, if the degree of relevance is greater than the predetermined threshold (YES in step S505), the related word search word and the word of the related word candidate are used as a related word pair, and the related word pair is associated with the related degree. Store in the associative dictionary storage unit 102 (step S506).
- the related word pairs for which the degree of relevance is newly calculated may be duplicated because one word may be a related word search word and another related word candidate word. And stored in the associative dictionary storage unit 102.
- the relevance is less than or equal to a predetermined threshold (NO in step S505), no processing is performed.
- steps S504 to step S406 are performed on all words included in the related word candidates (step S507), and this process (step S504 to step S507) is performed on all related word search words ( Step S50 8).
- the degree-of-association calculation unit 305 refers to the preference preference information stored in the preference preference information storage unit 303 and is selected by the word selection unit 203 (related word search word, (relationship Word candidates)) (Taro Matsushita, (Hanako Matsushita, Saburo Matsushita)), (Hanako Matsushita, (Taro Matsushita, Saburo Matsushita)), (Saburo Matsushita, (Taro Matsushita, Hanako Matsushita))))
- the degree of association between each related word search word and each word of the related word candidate is calculated.
- (related word search word, related word candidate word, relevance) are (Taro Matsushita, Hanako Matsushita, 60), (Taro Matsushita, Saburo Matsushita, 40), (Hanako Matsushita, Taro Matsushita, 60) , (Hanako Matsushita, Saburo Matsushita, 70), (Saburo Matsushita, Taro Matsushita, 40), (Saburo Matsushita, Hanako Matsushita, 70), and the degree of relevance is more than a predetermined threshold (50 as an example)
- a large (related word search word, related word candidate word, degree of association) is attenuated by multiplying the degree of association stored in the associative dictionary storage unit 102 shown in FIG. 2 by an attenuation parameter (for example, 0.8). This is added to the later associative dictionary storage unit 102.
- the association degree calculation unit 305 when updating the association dictionary storage unit 102, the association degree calculation unit 305 attenuates all association degrees stored in the association dictionary storage unit 102, and then newly If the same related word pair as the related word pair for which the relevance is calculated already exists in the associative dictionary storage unit 102, the relevance and the relevance calculation unit 305 already stored in the associative dictionary storage unit 102
- the power that is updated by adding the degree of relevance calculated in the above is not limited to this.
- the relevance calculation unit 305 may store the related word pairs for which the relevance is newly calculated in the associative dictionary storage unit 102 in association with the relevance.
- the moving image content is searched, but this is not a limitation, and it may be a hypertext or a web page on the Internet.
- the screen creation unit 205 stores the hypertext and web page display text searched by the search unit 110 in the display content storage unit 202, and further displays the searched noyper text and web page text. Convert to screen information and output the display screen information to the display unit 204.
- the screen creation unit 205 reads the file “name.html” from FIG. 11 (b).
- the display text “Taro Matsushita, Hanako Matsushita, Saburo Matsushita, Jiro Matsushita” is stored in the display content storage unit 202, and the file “name.html” is converted into display screen information. 11 Output display screen information as shown in (a). Furthermore, when all the text is not displayed on the display unit 204, the screen creation unit 205 may store only the displayed text in the display content storage unit 202. In the above embodiment, as an example, when the file “name.htmlj described in the HTML language of FIG. 10 is displayed as shown in FIG.
- the display content storage unit 202 may treat the text displayed on the display unit 204 for a predetermined time or more as the displayed text. This eliminates text that is only displayed when scrolling and is not viewed by the user.
- the screen creation unit 205 stores text information related to the content searched by the search unit 110 in the display content storage unit 202.
- the content is a video content and corresponds to each scene or frame of the video content. If there is text to be displayed, the screen creation unit 205 will only display the text corresponding to the scene or frame of the video content viewed by the user. May be stored in the display content storage unit 202. In this way, it is possible to create an associative dictionary that only uses text corresponding to scenes and frames of the video content that the user has watched. Can be created.
- the word stored in the associative dictionary is selected from the text related to the content viewed by the user, and the relevance level is determined for the content that matches the user's preference.
- the created association dictionary is used to make associations that match the user's preferences and to expand search conditions using words that are adapted to the user's viewing behavior. I can. Therefore, the expanded search result can be adapted to the user's viewing action, and the association can be adapted to the user's preference.
- the content storage unit 101 is a database storing information to be searched
- the word selection unit 105 is a word selection unit
- a relevance calculation unit 10 6 Corresponds to the related word storage means.
- the word selection unit 203 is a word selection unit
- the screen creation unit 205 is a display information extraction unit
- the relevance calculation unit 106 is a related word storage unit
- a search unit. 1 10 corresponds to search means.
- the preference content extraction unit 302 is a preference preference information extraction unit
- the word selection unit 304 is a word selection unit
- the relevance calculation unit 305 is a related word storage unit. Respectively.
- the preference content extraction unit 302 is a preference preference information extraction unit
- the word selection unit 203 is a word selection unit
- the relevance calculation unit 305 is a related word storage unit.
- the screen creation unit 205 corresponds to display information extraction means
- the search unit 110 corresponds to search means.
- the associative dictionary creation device adapts the search condition added for expansion when the search condition is expanded using the associative dictionary to the user's preference or viewing act. It is useful as a terminal that can search information by accessing the Internet, such as hard disk recorders, DVD recorders, TVs, audio components, and the Internet.
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Machine Translation (AREA)
Abstract
Description
Claims
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2006519685A JP3917648B2 (ja) | 2005-01-07 | 2005-12-27 | 連想辞書作成装置 |
US11/507,639 US8280720B2 (en) | 2005-01-07 | 2006-08-22 | Association dictionary creation apparatus |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2005-003171 | 2005-01-07 | ||
JP2005003171 | 2005-01-07 |
Related Child Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US11/507,639 Continuation US8280720B2 (en) | 2005-01-07 | 2006-08-22 | Association dictionary creation apparatus |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2006073095A1 true WO2006073095A1 (ja) | 2006-07-13 |
Family
ID=36647577
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2005/023919 WO2006073095A1 (ja) | 2005-01-07 | 2005-12-27 | 連想辞書作成装置 |
Country Status (4)
Country | Link |
---|---|
US (1) | US8280720B2 (ja) |
JP (1) | JP3917648B2 (ja) |
CN (1) | CN100454303C (ja) |
WO (1) | WO2006073095A1 (ja) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2008056570A1 (fr) * | 2006-11-09 | 2008-05-15 | Panasonic Corporation | Dispositif de recherche de contenu |
JP2009069874A (ja) * | 2007-09-10 | 2009-04-02 | Sharp Corp | コンテンツ検索装置、コンテンツ検索方法、プログラム、および記録媒体 |
JP2012008900A (ja) * | 2010-06-28 | 2012-01-12 | Nippon Telegr & Teleph Corp <Ntt> | 関連語句抽出方法、関連語句抽出装置及び関連語句抽出プログラム |
US8504357B2 (en) | 2007-08-03 | 2013-08-06 | Panasonic Corporation | Related word presentation device |
Families Citing this family (20)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8793569B2 (en) * | 2005-04-01 | 2014-07-29 | Sony Corporation | Presenting a recommendation based on user preference |
JP2007206975A (ja) * | 2006-02-01 | 2007-08-16 | Toshiba Corp | 言語情報変換装置及びその方法 |
JP2007241888A (ja) * | 2006-03-10 | 2007-09-20 | Sony Corp | 情報処理装置および方法、並びにプログラム |
US8037086B1 (en) | 2007-07-10 | 2011-10-11 | Google Inc. | Identifying common co-occurring elements in lists |
JP2009025968A (ja) * | 2007-07-18 | 2009-02-05 | Fujifilm Corp | 関連語辞書作成装置、方法、及びプログラム、並びにコンテンツ検索装置 |
JP2009302884A (ja) * | 2008-06-13 | 2009-12-24 | Sony Corp | 情報処理装置、情報処理方法、およびプログラム |
CN102411563B (zh) * | 2010-09-26 | 2015-06-17 | 阿里巴巴集团控股有限公司 | 一种识别目标词的方法、装置及系统 |
CN102479236B (zh) * | 2010-11-30 | 2015-09-23 | 成都致远诺亚舟教育科技有限公司 | 一种单词关联库系统及其实现方法和电子学习设备 |
CN102567423B (zh) * | 2010-12-31 | 2014-12-03 | 成都致远诺亚舟教育科技有限公司 | 一种诗词关联搜索方法和系统 |
CN102567424B (zh) * | 2010-12-31 | 2014-12-03 | 成都致远诺亚舟教育科技有限公司 | 一种诗词关联库系统及其实现方法和电子学习设备 |
CN102622371B (zh) * | 2011-01-28 | 2015-06-03 | 成都致远诺亚舟教育科技有限公司 | 一种历史关联库系统及其实现方法和电子学习设备 |
US9785628B2 (en) * | 2011-09-29 | 2017-10-10 | Microsoft Technology Licensing, Llc | System, method and computer-readable storage device for providing cloud-based shared vocabulary/typing history for efficient social communication |
TWI456412B (zh) * | 2011-10-11 | 2014-10-11 | Univ Ming Chuan | 製作知識地圖的方法 |
JP6342678B2 (ja) * | 2014-03-07 | 2018-06-13 | クラリオン株式会社 | 関連データ生成装置、関連データ生成方法およびプログラム |
CN105653533B (zh) * | 2014-11-13 | 2019-10-25 | 腾讯数码(深圳)有限公司 | 一种更新分类关联词集合的方法和装置 |
US9734141B2 (en) | 2015-09-22 | 2017-08-15 | Yang Chang | Word mapping |
US11189193B2 (en) | 2016-04-26 | 2021-11-30 | Ponddy Education Inc. | Affinity knowledge based computational learning system |
CN110032675A (zh) * | 2019-03-13 | 2019-07-19 | 平安城市建设科技(深圳)有限公司 | 基于共现词的检索方法、装置、设备及可读存储介质 |
CN110059156A (zh) * | 2019-03-13 | 2019-07-26 | 平安城市建设科技(深圳)有限公司 | 基于关联词的协同检索方法、装置、设备及可读存储介质 |
CN110069599A (zh) * | 2019-03-13 | 2019-07-30 | 平安城市建设科技(深圳)有限公司 | 基于近似词的检索方法、装置、设备及可读存储介质 |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2001117935A (ja) * | 1999-10-19 | 2001-04-27 | Fujitsu Ltd | テキスト情報アクセス装置、その方法、及び記録媒体 |
Family Cites Families (22)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH0877203A (ja) | 1994-09-07 | 1996-03-22 | Fuji Xerox Co Ltd | 文書検索装置 |
JP3625344B2 (ja) * | 1996-11-05 | 2005-03-02 | 株式会社ビデオリサーチ | 視聴チャンネル検出装置 |
JP3099756B2 (ja) | 1996-10-31 | 2000-10-16 | 富士ゼロックス株式会社 | 文書処理装置、単語抽出装置及び単語抽出方法 |
JP3596210B2 (ja) | 1997-01-16 | 2004-12-02 | 富士ゼロックス株式会社 | 関連語辞書作成装置 |
US6029165A (en) * | 1997-11-12 | 2000-02-22 | Arthur Andersen Llp | Search and retrieval information system and method |
US6553566B1 (en) * | 1998-08-27 | 2003-04-22 | X Out Corporation | Viewer controlled multi-function system for processing television signals |
US7209942B1 (en) * | 1998-12-28 | 2007-04-24 | Kabushiki Kaisha Toshiba | Information providing method and apparatus, and information reception apparatus |
JP2000348042A (ja) * | 1999-06-03 | 2000-12-15 | Fujitsu Ltd | 統合シソーラス作成装置、修正シソーラス作成装置、情報収集型シソーラス作成装置、統合シソーラス作成プログラム記憶媒体、修正シソーラス作成プログラム記憶媒体、および情報収集型シソーラス作成プログラム記憶媒体 |
US7181438B1 (en) * | 1999-07-21 | 2007-02-20 | Alberti Anemometer, Llc | Database access system |
JP3994368B2 (ja) * | 2000-01-25 | 2007-10-17 | ソニー株式会社 | 情報処理装置および情報処理方法、並びに記録媒体 |
US20020059204A1 (en) * | 2000-07-28 | 2002-05-16 | Harris Larry R. | Distributed search system and method |
JP2002077755A (ja) * | 2000-08-29 | 2002-03-15 | Sharp Corp | エージェントインタフェース装置 |
JP2002082987A (ja) * | 2000-09-06 | 2002-03-22 | Seiko Epson Corp | 文書情報閲読支援装置、ディジタルコンテンツ作成システム、ディジタルコンテンツ配信システム及び記憶媒体 |
US7065550B2 (en) * | 2001-02-14 | 2006-06-20 | International Business Machines Corporation | Information provision over a network based on a user's profile |
US7191115B2 (en) * | 2001-06-20 | 2007-03-13 | Microsoft Corporation | Statistical method and apparatus for learning translation relationships among words |
US20030126227A1 (en) * | 2001-12-31 | 2003-07-03 | Koninklijke Philips Electronics N.V | Method of populating an explicit profile |
US7640164B2 (en) * | 2002-07-04 | 2009-12-29 | Denso Corporation | System for performing interactive dialog |
US7657907B2 (en) * | 2002-09-30 | 2010-02-02 | Sharp Laboratories Of America, Inc. | Automatic user profiling |
US7340460B1 (en) * | 2003-04-04 | 2008-03-04 | Yahoo! Inc. | Vector analysis of histograms for units of a concept network in search query processing |
JP2004348607A (ja) | 2003-05-23 | 2004-12-09 | Nippon Telegr & Teleph Corp <Ntt> | コンテンツ検索方法、コンテンツ検索システム、コンテンツ検索用プログラムおよびコンテンツ検索用プログラムが記録された記録媒体 |
CA2532257A1 (en) * | 2003-07-14 | 2005-01-27 | Orative Corporation | System and method for active mobile collaboration |
JP4008954B2 (ja) * | 2004-10-29 | 2007-11-14 | 松下電器産業株式会社 | 情報検索装置 |
-
2005
- 2005-12-27 JP JP2006519685A patent/JP3917648B2/ja active Active
- 2005-12-27 CN CNB2005800081455A patent/CN100454303C/zh active Active
- 2005-12-27 WO PCT/JP2005/023919 patent/WO2006073095A1/ja not_active Application Discontinuation
-
2006
- 2006-08-22 US US11/507,639 patent/US8280720B2/en active Active
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2001117935A (ja) * | 1999-10-19 | 2001-04-27 | Fujitsu Ltd | テキスト情報アクセス装置、その方法、及び記録媒体 |
Non-Patent Citations (1)
Title |
---|
SHIMOHATA M.: "Renso Jisho o Riyo shita Imi Kensaku System", OKIDENKI KENKYU KAIHATSU, vol. 65, no. 1, 1 January 1998 (1998-01-01), pages 15 - 18, XP002934270 * |
Cited By (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2008056570A1 (fr) * | 2006-11-09 | 2008-05-15 | Panasonic Corporation | Dispositif de recherche de contenu |
US7904449B2 (en) | 2006-11-09 | 2011-03-08 | Panasonic Corporation | Content searching apparatus |
US8504357B2 (en) | 2007-08-03 | 2013-08-06 | Panasonic Corporation | Related word presentation device |
JP2009069874A (ja) * | 2007-09-10 | 2009-04-02 | Sharp Corp | コンテンツ検索装置、コンテンツ検索方法、プログラム、および記録媒体 |
JP2012008900A (ja) * | 2010-06-28 | 2012-01-12 | Nippon Telegr & Teleph Corp <Ntt> | 関連語句抽出方法、関連語句抽出装置及び関連語句抽出プログラム |
Also Published As
Publication number | Publication date |
---|---|
CN1930567A (zh) | 2007-03-14 |
JP3917648B2 (ja) | 2007-05-23 |
JPWO2006073095A1 (ja) | 2008-06-12 |
CN100454303C (zh) | 2009-01-21 |
US8280720B2 (en) | 2012-10-02 |
US20060282258A1 (en) | 2006-12-14 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2006073095A1 (ja) | 連想辞書作成装置 | |
JP4008954B2 (ja) | 情報検索装置 | |
US8843467B2 (en) | Method and system for providing relevant information to a user of a device in a local network | |
US8782056B2 (en) | Method and system for facilitating information searching on electronic devices | |
US7725467B2 (en) | Information search system, information processing apparatus and method, and information search apparatus and method | |
US6569206B1 (en) | Facilitation of hypervideo by automatic IR techniques in response to user requests | |
US20080183681A1 (en) | Method and system for facilitating information searching on electronic devices | |
KR102154735B1 (ko) | 프로그램 추천 장치 및 프로그램 추천 프로그램 | |
JP4399514B2 (ja) | 対話型番組検索装置 | |
JP4487018B2 (ja) | 関連シーン付与装置及び関連シーン付与方法 | |
JP2005115790A (ja) | 情報検索方法、情報表示装置及びプログラム | |
CN1326075C (zh) | 自动视频检索器精灵 | |
JP2010097249A (ja) | 電子機器、コンテンツ推薦方法及びプログラム | |
US20090055393A1 (en) | Method and system for facilitating information searching on electronic devices based on metadata information | |
JP2009043156A (ja) | 番組検索装置および番組検索方法 | |
JPH1069496A (ja) | インターネット検索装置 | |
JP2009266096A (ja) | 情報処理装置、および関連アイテムの提供方法 | |
JP5491372B2 (ja) | 情報検索システム、情報検索方法、情報検索プログラム | |
JP5242105B2 (ja) | 情報処理装置および情報表示方法 | |
JP2010218385A (ja) | コンテンツ検索装置及びコンピュータプログラム | |
JP2004362121A (ja) | 情報検索装置、携帯情報端末装置、情報検索方法、情報検索プログラム及び記録媒体 | |
WO2011067808A1 (ja) | 番組推薦装置および番組推薦方法 | |
JP4477931B2 (ja) | 検索リクエスト装置、検索リクエスト方法、検索リクエストプログラムおよび検索リクエストプログラムを記録したコンピュータ読み取り可能な記録媒体 | |
JP2001101195A (ja) | テレビ番組検索装置および自動番組選択装置 | |
JP3759700B2 (ja) | 利用者嗜好反映型検索方法及びその実施装置並びにその処理プログラムと記録媒体 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
WWE | Wipo information: entry into national phase |
Ref document number: 2006519685 Country of ref document: JP |
|
WWE | Wipo information: entry into national phase |
Ref document number: 11507639 Country of ref document: US |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application | ||
WWE | Wipo information: entry into national phase |
Ref document number: 200580008145.5 Country of ref document: CN |
|
WWP | Wipo information: published in national office |
Ref document number: 11507639 Country of ref document: US |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 05822386 Country of ref document: EP Kind code of ref document: A1 |
|
WWW | Wipo information: withdrawn in national office |
Ref document number: 5822386 Country of ref document: EP |