WO2012063770A1 - 関連語登録装置、情報処理装置、関連語登録方法、関連語登録装置用プログラム、および、記録媒体 - Google Patents
関連語登録装置、情報処理装置、関連語登録方法、関連語登録装置用プログラム、および、記録媒体 Download PDFInfo
- Publication number
- WO2012063770A1 WO2012063770A1 PCT/JP2011/075572 JP2011075572W WO2012063770A1 WO 2012063770 A1 WO2012063770 A1 WO 2012063770A1 JP 2011075572 W JP2011075572 W JP 2011075572W WO 2012063770 A1 WO2012063770 A1 WO 2012063770A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- character string
- related word
- search
- word
- search query
- Prior art date
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/24—Querying
- G06F16/245—Query processing
- G06F16/2453—Query optimisation
-
- 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
- G06F16/33—Querying
- G06F16/332—Query formulation
- G06F16/3322—Query formulation using system suggestions
-
- 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
- G06F16/374—Thesaurus
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/40—Data acquisition and logging
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/10—Text processing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/237—Lexical tools
- G06F40/247—Thesauruses; Synonyms
Definitions
- the present invention relates to a technical field of a related word registration device, an information processing device, a related word registration method, a related word registration device program, and a recording medium for registering a search keyword as a related word.
- Patent Literature 1 discloses a semantic dictionary registration device that automatically assigns semantic information to an input dictionary entry using information on a dictionary entry that has already been given semantic information in a semantic dictionary. .
- Patent Literature 1 since the similarity is determined based on the semantic distance between simple words of the input data and all registered data, the accuracy of the determination of the similarity is low. , Registration of related terms such as unnecessary synonyms could not be suppressed. Therefore, it has been difficult to construct a dictionary of related terms such as a synonym dictionary with high accuracy.
- the present invention has been made in view of such problems, and an object of the present invention is to provide a related word registration device or the like that can register related words with high accuracy.
- the invention according to claim 1 is related to generating a related word candidate set by extracting related word candidates from a search query log storing a log related to a search query based on a predetermined condition.
- Word candidate set generation means, reception means for receiving a search query for a search word input by a user, partial character string generation means for generating a partial character string from a character string of the search word, and the generated partial character string A candidate character string extracting unit that extracts a candidate character string from the related word candidate set, a similarity between the candidate character string and the search word, the number of uses of the candidate character string, and the candidate character string
- a criterion generation unit that generates a criterion line of the suitability score for the ranking as a criterion for determining whether or not to
- a registered character string extracting means for extracting a candidate character string whose deviation between the suitability score and the reference line is a preset threshold value or more as a registered character string, and the extracted registered character string;
- a related word registration means for registering the search word as a related word in a related word database.
- the partial character string generation unit sets the length of the partial character string based on the number of characters of the search word, and the setting is performed. It generates a partial character string of a certain length.
- the partial character string generation means determines the number of characters of the search word when generating the partial character string based on the number of characters of the search word.
- a moving width is set, and a partial character string is generated with the set moving width.
- the partial character string generation means converts the search word into a search word that has been phonetically converted. On the other hand, the partial character string is generated.
- the parameter adjustment is performed to adjust the similarity and the information on the number of times of use of the score calculation means. Means are further provided.
- the determination criterion generating unit approximates a relationship between the ranking and the suitability score.
- the reference line is obtained from a function or from the ranking data within a predetermined range.
- the registered character string extracting means sets an upper limit on the number of the registered character strings to be extracted. It is characterized by that.
- the invention according to claim 8 stores the search word related to the product sold on the internal or external EC site in the related word registration device according to any one of claims 1 to 7.
- a search word storage means is further provided.
- the score calculation means calculates the suitability score based on log data related to the search query. It is characterized by calculating.
- a tenth aspect of the present invention is the related word registration device according to any one of the first to ninth aspects, wherein the received search query is stored in accordance with the reception order, and preset.
- the search query extraction means for extracting the search query earlier in the receiving order than the received search query from the search query storage means, and the destination constituting the extracted search query
- Character string set storage means for storing the search word of the received search query and the search word constituting the received search query as a character string set, and the character string set storage means in accordance with a preset character string set extraction start condition
- a character string extracting means for extracting a character string set whose search word is the same or similar from the above, and a related word from the extracted character string set based on a preset registration condition Becomes a related word specifying means for specifying a character string set, further wherein the related-word registration unit, and registers the associated word database the specific character string set as a related word.
- the invention according to claim 11 is the related word registration device according to claim 10, wherein the search query storage means further stores user identification information, and the search query extraction means has the search query extraction condition as: The previous search query that matches the user identification information of the user who has input the search query is extracted.
- the search query extraction unit includes, as the search query extraction condition, within a predetermined time from the reception time of the search query. In this case, a search query received in advance is extracted.
- the search query storage means further stores category information
- the search query extraction means The previous search query is extracted based on the category information as the search query extraction condition.
- the character string extraction unit uses the search query storage unit as the character string set extraction start condition.
- the search query storage unit uses the search query storage unit as the character string set extraction start condition.
- the character string extraction unit satisfies a predetermined condition after the character string set extraction start condition is satisfied.
- the character string set is extracted when time elapses.
- the related word specifying unit is configured such that the previous search word is the same as the registration condition.
- the character It is characterized by specifying a column set as a related word.
- the invention according to claim 17 is an information processing apparatus for performing information processing with reference to the related word database of the related word registration apparatus according to any one of claims 1 to 16, wherein the related word database And a related word extracting unit that extracts a related word corresponding to a search word of the received search query, and a related word output unit that outputs a related word extracted by the related word extracting unit. It is characterized by that.
- a related word candidate is extracted based on a predetermined condition from a search query log storing a log related to a search query.
- a ranking step for ranking the candidate character strings in sequence, and a criterion for a suitability score for the ranking as a criterion for determining whether or not to register the candidate character string as a related word based on the suitability score and ranking of the candidate character string A criterion generation step for generating a line, and a registered character string extraction step for extracting a candidate character string
- related word candidate set generation means for generating a related word candidate set by extracting related word candidates from a search query log storing a log related to a search query based on a predetermined condition.
- candidate character string extraction means for extracting candidate character strings from the set, the similarity between the candidate character string and the search word, the number of uses of the candidate character string, and the number of search results of the search using the candidate character string
- Score calculating means for calculating suitability scores of the candidate character strings, ranking means for ranking the candidate character strings in the order of the scores, and the candidate sentences
- Judgment criterion generating means for generating a reference line of suitability scores for the ranking as a criterion for determining whether or not to register the candidate character string as a related
- the invention according to claim 20 is a related word candidate set generation means for generating a related word candidate set by extracting a related word candidate from a search query log storing a log related to a search query based on a predetermined condition.
- candidate character string extraction means for extracting candidate character strings from the set, the similarity between the candidate character string and the search word, the number of uses of the candidate character string, and the number of search results of the search using the candidate character string
- Judgment criterion generating means for generating a reference line of suitability scores for the ranking as a criterion for determining whether or not to register the candidate character string as
- a related word candidate set obtained by extracting related word candidates based on a predetermined condition is generated from a search query log storing a log related to a search query, and a search word search query input by a user is received.
- Generate a partial character string from the character string of the search word extract the candidate character string from the related word candidate set based on the partial character string, the similarity between the candidate character string and the search word, the number of times the candidate character string is used, And based on the number of search results of the search with the candidate character string, the suitability score of the candidate character string is calculated, the candidate character strings are ranked in order of the score, and the candidate based on the suitability score and the ranking of the candidate character string
- a reference line of suitability scores for ranking is generated as a criterion for determining whether or not to register a character string as a related word, and the deviation between the suitability score and the reference line is preset.
- the number of characters is reduced by extracting candidate character strings that are equal to or higher than the threshold as registered character strings for registering as related words, and registering the extracted registered character strings and search words as related words in the related word database. Since a candidate character string is extracted from the related word candidate set based on the partial character string thus formed, registration of unnecessary related words can be suppressed by the suitability score of the candidate character string, and the accuracy of the related words can be improved.
- FIG. 6 is a flowchart showing an operation example of generation of a related word candidate set of a related word registration server in the operation of the related word registration system 1 of FIG. 1 according to the first embodiment. It is a schematic diagram which shows an example of the web page displayed on the terminal of FIG.
- FIG. 1 is a schematic diagram showing a schematic configuration example of a related word registration system 1 according to the present embodiment.
- a related word registration system 1 includes a related word registration server 10 (an example of a related word registration device) for registering related words, and a shopping site (internal EC) that is an example of an information providing site used by a user. (Example of electronic commerce) site)
- An information providing server 20 for providing users with information and product information that is installed for management and registered in the related word registration server 10, and the information providing server 20 And a terminal 30 for searching for products and the like.
- the related word registration server 10 and the information providing server 20 are connected by a local area network or the like, and are capable of transmitting and receiving data to each other, and constitute a server system 5.
- the server system 5 and the terminal 30 are connected by the network 3 and can transmit and receive data using a communication protocol (for example, TCP / IP).
- the network 3 is constructed by, for example, the Internet, a dedicated communication line (for example, a CATV (Community Antenna Television) line), a mobile communication network (including a base station, etc.), a gateway, and the like.
- the related word registration server 10 builds a database of related words from the search behavior of the user in the terminal 30 or the like, and supports the search behavior of the user in the terminal 30 or the like.
- the information providing server 20 receives a search query including a search word from the terminal 30 and performs a search, and transmits the search result to the terminal 30. In addition, the information providing server 20 transmits the received search query to the related word registration server 10 in order to construct a related word database. In addition, the information providing server 20 performs a product search, product information provision, advertisement information provision, product purchase procedure, and the like as a shopping site.
- FIG. 2 is a block diagram illustrating an example of a schematic configuration of the related word registration server 10.
- the related word registration server 10 that functions as a computer includes a communication unit 11, a storage unit 12, an input / output interface unit 13, and a system control unit 14.
- the system control unit 14 and the input / output interface unit 13 are connected via a system bus 15.
- the communication unit 11 is connected to the network 3 to control the communication state with the terminal 30 and the like, and is further connected to the local area network to transmit and receive data to and from other servers such as the information providing server 20 on the local area network. I do.
- the storage unit 12 is configured by, for example, a hard disk drive or the like, and stores various programs such as an operating system and a server program, data, and the like. Note that the various programs may be acquired from, for example, another server device via the network 3, or may be recorded on a recording medium and read via a drive device (not shown). Also good.
- the storage unit 12 also includes a search query log database 12a (hereinafter referred to as “search query log DB 12a”) that stores a log related to the search query received from the terminal 30, and a candidate character string of related words generated from the search query.
- search query log DB 12a search query log database 12a
- Related word candidate database 12b (hereinafter referred to as “related word candidate DB 12b") stored as a related word candidate set
- related word database 12c hereinafter referred to as “related word DB 12c”
- character string set database 12d hereinafter referred to as “character string set DB 12d” and the like.
- the search query log DB 12a, the related word candidate DB 12b, the related word DB 12c, and the character string set DB 12d may be constructed in a server other than the related word registration server 10 in the server system 5, or the server system 5 It may be built externally.
- the search query log DB 12a (an example of a search query storage unit) stores a search query associated with user identification information for identifying a user such as a reception time and a user ID when the search query is received.
- the user identification information includes a user ID of a shopping site, a user terminal or access point identification number, an IP address, and the like.
- the search query is also associated with category information such as the product category to which the product belongs and the product category indicated by the Web page opened by the user. It is stored in the search query log DB 12a.
- related word candidate DB 12b candidate character strings of related words extracted based on predetermined conditions from the search query log of the search query log DB 12a are stored as a related word candidate set.
- related words generated by the related word registration server 10 are stored.
- the character string set DB 12d (an example of a character string set storage unit) stores a character string set of the search word of the received search query and the search word of the search query extracted from the search query log DB 12a.
- the input / output interface unit 13 performs interface processing between the communication unit 11 and the storage unit 12 and the system control unit 14.
- the system control unit 14 includes a CPU (Central Processing Unit) 14a, a ROM (Read Only Memory) 14b, a RAM (Random Access Memory) 14c, and the like.
- the CPU 14 a reads and executes various programs stored in the ROM 14 b and the storage unit 12 to perform related word registration processing and the like.
- FIG. 3 is a block diagram illustrating an example of a schematic configuration of the information providing server 20.
- the information providing server 20 includes a communication unit 21, a storage unit 22, an input / output interface unit 23, and a system control unit 24, and the system control unit 24, the input / output interface unit 23, and the like. Are connected via a system bus 25.
- the configuration and function of the information providing server 20 are substantially the same as the configuration and function of the related word registration server 10, and therefore, different points will be mainly described in each configuration and function of the related word registration server 10.
- the communication unit 21 controls the communication state with the terminal 30 and the related word registration server 10 through the network 3 and the local area network.
- product DB product database
- member DB member database
- the product DB 22a is associated with a product ID that is an identifier for identifying a product, and stores a product name, type, product image, specifications, product information, advertisement information about each product, and the like.
- the product DB 22a stores a product web page file described in a markup language such as HTML (HyperText Markup Language), XML (Extensible ⁇ Markup Language), or the like.
- the product DB 22a is an example of a search word storage unit that stores a search word related to a product sold on an EC site inside the server system 5.
- the user ID, name, address, telephone number, e-mail address, occupation, hobby, purchase history, theme and genre of interest of the user (product category) of the registered user (shopping site user) are stored in the member DB 22b. Such user information is registered.
- a user ID, a login ID, and a password necessary for the user to log in to the shopping site from the terminal 30 are registered.
- the login ID and the password are login information used for login processing (user authentication processing).
- the system control unit 24 includes a CPU 24a, a ROM 24b, a RAM 24c, and the like.
- the CPU 24a reads out and executes various programs stored in the ROM 24b and the storage unit 22, thereby performing product search processing, product purchase processing by the user, and the like.
- FIG. 4 is a block diagram illustrating an example of a schematic configuration of the terminal 30.
- the terminal 30 that functions as a computer is a portable terminal such as a portable wireless telephone or a PDA including a personal computer or a smartphone, and includes a communication unit 31, a storage unit 32, a display unit 33, and the like. , An operation unit 34, an input / output interface unit 35, and a system control unit 36.
- the system control unit 36 and the input / output interface unit 35 are connected via a system bus 37.
- the communication unit 31 controls communication with the information providing server 20 and the like through the network 3.
- the communication unit 31 has a wireless communication function in order to connect to the mobile communication network of the network 3.
- the storage unit 32 includes, for example, a hard disk drive and stores an operating system, a web browser program, a web browser toolbar program, and the like.
- the display unit 33 is configured by, for example, a liquid crystal display element or an EL (Electro Luminescence) element.
- a liquid crystal display element or an EL (Electro Luminescence) element.
- the web page of the search screen and the web page of the result of the product search provided from the information providing server 20 are displayed by the web browser.
- the operation unit 34 includes, for example, a keyboard and a mouse. The user inputs a response through the operation unit 34.
- the display unit 33 is a touch switch type display panel such as a touch panel
- the operation unit 34 acquires position information of the display unit 33 that is touched or approached by the user.
- the input / output interface unit 35 is an interface between the communication unit 31 and the storage unit 32 and the system control unit 36.
- the system control unit 36 includes, for example, a CPU 36a, a ROM 36b, and a RAM 36c.
- the CPU 36a reads out and executes various programs stored in the ROM 36b, the RAM 36c, and the storage unit 32.
- the system control unit 36 executes a web browser program and functions as a web browser.
- FIG. 5 is a flowchart showing an operation example of generation of a related word candidate set of the related word registration server 10.
- FIG. 6 is a schematic diagram illustrating an example of a web page displayed on the terminal 30.
- FIG. 7 is a flowchart showing an operation example of registration of related words in the related word registration server.
- FIG. 8 is a schematic diagram illustrating an example of partial character string extraction.
- FIG. 9 is a schematic diagram showing an example of a ranking-score graph.
- FIG. 10 is a flowchart showing a subroutine of an operation example for calculating the suitability score in the related word registration server 10.
- FIG. 11 is a diagram showing an example of an availability pattern for calculating a suitability score.
- the information providing server 20 when the information providing server 20 receives a search query including a search word input by the user from the terminal 30, the information providing server 20 performs a search and transmits the search query to the related word registration server 10. Further, the information providing server 20 also transmits the number of search results for the search query to the related word registration server 10.
- the related word registration server 10 receives the search query and the number of search results from the information providing server 20, and stores the search query in the search query log DB 12a in association with the number of search results. At this time, the related word registration server 10 may store the search query in association with the user identification information for distinguishing the user such as the reception time when the search query is received, the user ID and the IP address.
- the related word registration server 10 extracts a search query from the search query log (step S1). Specifically, the system control unit 14 of the related word registration server 10 extracts one search query from the search query log DB 12a.
- the related word registration server 10 determines whether or not the number of search results is one or more search queries (step S2). Specifically, the system control unit 14 of the related word registration server 10 reads the number of search results corresponding to the extracted search query from the search query log DB 12a, and determines whether the number of search results is one or more. To do. The system control unit 14 of the related word registration server 10 transmits the extracted search query to the information providing server 20 and receives the number of hits from the information providing server 20 in order to obtain the number of search results hit by the search query. You may comprise as follows.
- the related word registration server 10 registers the extracted search query in the related word candidate set (step S3). Specifically, the system control unit 14 of the related word registration server 10 determines the search query (related word candidate) when the number of search results is one or more, that is, the number of search results is not zero. (Including a search word as a character string) is registered in the related word candidate DB 12b as a related word candidate set.
- step S2 If the number of search results is not one or more (step S2; NO), the related word registration server 10 does not register the extracted search query in the related word candidate set.
- step S4 it is determined whether or not there are remaining search queries. Specifically, the system control unit 14 of the related word registration server 10 determines whether there is a search query that has not yet been determined for the number of search results in the search query log DB 12a.
- step S4 When there is a remaining search query (step S4; YES), the process returns to step S1, and the related word registration server 10 extracts the next search query, and when there is no remaining search query (step S4; NO), The process of generating the related word candidate set is terminated.
- the related word registration server 10 generates a related word candidate set of a certain scale in advance.
- the related word registration server 10 functions as an example of a related word candidate set generation unit that generates a related word candidate set by extracting related word candidates based on a predetermined condition from a search query log storing a log related to a search query. .
- the information providing server 20 When the information providing server 20 receives a search query and performs a search, if the number of search results is not zero, the information providing server 20 adds information that is not zero to the related query registration server 10. You may send it. In this case, when the related word registration server 10 receives a search query in which the number of search results is not zero, the related word registration server 10 stores it in the related word candidate DB 12b.
- the search word input field 41 of the Web page 40 provided from the information providing server 20 is displayed on the display unit 33 of the terminal 30 by, for example, “A family table” and the search word from the user. Entered. A search is performed by clicking the search button 42, and the terminal 30 transmits a search query including this search word to the information providing server 20.
- the information providing server 20 receives a search query including a search word input by the user from the terminal 30, the information providing server 20 transmits the search query to the related word registration server 10.
- the related word registration server 10 receives a search query including a search word input by the user (step S10). Specifically, the system control unit 14 of the related word registration server 10 receives a search query from the information providing server 20 through the communication unit 11. Note that, as described in the generation of the related word candidate set, the related word registration server 10 stores the received search query in the search query log DB 12a. Moreover, the related word registration server 10 may register a search query in a related word candidate set, when the conditions of step S2 are satisfy
- the related word registration server 10 acquires the number of characters of the search word (step S11). Specifically, when the search word included in the search query is kanji or the like, the system control unit 14 of the related word registration server 10 converts it into a reading such as hiragana or converts it into romaji. For example, the system control unit 14 of the related word registration server 10 converts the phonetic table of “A family's table” to “excellent” and increases the number of characters from 5 to 9 characters. Furthermore, the system control unit 14 of the related word registration server 10 may convert “eike no shokutaku” into “eikenoshokutaku” in Roman letters and increase it to 15 characters. Then, the system control unit 14 of the related word registration server 10 acquires the number of characters (for example, 9 characters or 15 characters) of the search word subjected to phonetic conversion such as hiragana conversion or romaji conversion.
- the related word registration server 10 determines whether or not the number of characters in the search word is 5 or more (step S12). Specifically, the system control unit 14 of the related word registration server 10 determines whether or not the number of characters of the search word subjected to phonetic conversion is five or more.
- the related word registration server 10 extracts a 4-character partial character string from the search word (step S13). Specifically, as shown in FIG. 8, the system control unit 14 of the related word registration server 10 sets the length (Window Size) L of the partial character string to 4 for the character string “eikenoshokutaku” of the search word. Then, the sliding width (Sliding Size) S of the window is set to 2, and partial character strings such as the partial character string p1 (“eike”) and the partial character string p2 (“keno”) are generated and extracted. Note that the threshold value (5 characters) in step S2 is set to a value larger than the length L of the partial character string based on the length L of the partial character string.
- the related word registration server 10 functions as an example of a partial character string generation unit that generates a partial character string from a character string of a search word.
- the related word registration server 10 functions as an example of a partial character string generation unit that sets the length of the partial character string based on the number of characters of the search word and generates the partial character string having the set length. Further, the related word registration server 10 sets the movement width of the number of characters of the search word when generating the partial character string based on the number of characters of the search word, and generates the partial character string with the set movement width. It functions as an example of a column generation unit.
- the related word registration server 10 functions as an example of a partial character string generation unit that generates a partial character string for a search word obtained by phonetic conversion of the search word.
- the partial match is a candidate character string including “eike” or a candidate character string including a part “eik” of “eike”.
- the related word registration server 10 functions as an example of a candidate character string extraction unit that extracts a candidate character string from a related word candidate set based on the generated partial character string.
- the related word registration server 10 extracts a related character candidate character string partially matching the search word from the related word candidate set (Ste S15). Specifically, the system control unit 14 of the related word registration server 10 does not extract the partial character string, but selects the received search word itself or a candidate character string that partially matches the phonetic converted search word as the related word candidate DB 12b. Extract from
- system control unit 14 of the related word registration server 10 may extract a candidate character string similar to a partial character string or a search word based on the similarity.
- the similarity includes distances between character strings such as Jaro-Winker distance and Levenstein distance.
- the related word registration server 10 calculates a suitability score for each related word candidate character string (step S16). Specifically, the system control unit 14 of the related word registration server 10 calculates a suitability score for each extracted candidate character string by a suitability score subroutine described later.
- the related word registration server 10 obtains rankings in order of suitability scores (step S17). Specifically, the system control unit 14 of the related word registration server 10 performs a process of arranging the candidate character strings in descending order of suitability scores. Then, the system control unit 14 of the related word registration server 10 takes out a certain upper character string, for example, the upper 50 candidate character strings, and further narrows down the candidate character strings. As described above, the related word registration server 10 functions as an example of a ranking unit that ranks candidate character strings in the order of scores.
- the related word registration server 10 generates a reference line of suitability scores (step S18). Specifically, the system control unit 14 of the related word registration server 10 determines and generates a reference line from words having a higher suitability score among the extracted upper related word candidate character strings. More specifically, as shown in FIG. 9A, in the graph illustrating the relative suitability score with respect to the ranking, the system control unit 14 approximates the points of the suitability score from the 1st place to the nth place, for example. Let the function be the reference line L1.
- an approximation function as an example of the reference line, a logarithmic function, an exponential function, a linear function, a power function such as a quadratic function or a cubic function, a function of a conic curve such as a hyperbola, or the like can be given.
- the approximation function may be a function that approximates or interpolates each point of the ranking-score graph.
- a criterion such as a least square method may be used.
- the relative suitability score is, for example, a value obtained by dividing each score by the value of the suitability score ranked first.
- the lower suitability score and the scores from the 50th to the 10th may be determined by the least square method. Since the ranking-score curve is arranged in the ranking of the suitability score, it becomes a monotonically decreasing graph in a broad sense.
- the related word registration server 10 generates a reference line of the suitability score for the ranking as a reference for determining whether or not to register the candidate character string as a related word based on the suitability score and ranking of the candidate character string. It functions as an example of means.
- the related word registration server 10 functions as an example of a determination criterion generation unit that obtains a reference line from an approximation function that approximates the relationship between the ranking and the suitability score.
- the related word registration server 10 functions as an example of a criterion generation unit that obtains a reference line from ranking data in a predetermined range.
- the related word registration server 10 determines whether or not the difference between each suitability score and the reference line is greater than or equal to a threshold value (step S19). Specifically, the system control unit 14 of the related word registration server 10 determines whether or not the difference between the reference line L1 and the first ranking score is equal to or greater than a threshold value. More specifically, the system control unit 14 of the related word registration server 10 calculates a difference by subtracting a value obtained by assigning the ranking value to the formula of the reference line L1 from the suitability score of the first ranking. If the difference is greater than or equal to the threshold ⁇ , the candidate character string is extracted as a related character registration character string.
- the threshold value is obtained in advance by simulation, for example, 0.1 times the reference line by changing a parameter or the like for adjusting the suitability score.
- the system control unit 14 of the related word registration server 10 extracts candidate character strings whose suitability scores are 1.1 times or more of the reference line as related words.
- the related word registration server 10 extracts a registered character string that extracts a candidate character string whose deviation between the suitability score and the reference line L1 is equal to or greater than a preset threshold value ⁇ as a related character string. It functions as an example of means.
- the related word registration server 10 registers a candidate character string of the related word whose deviation from the reference line is equal to or larger than the threshold as a related word (Ste S20). Specifically, the system control unit 14 of the related word registration server 10 registers the related word of the received search word as the candidate character string ranked first when the deviation between each suitability score and the reference line is equal to or greater than the threshold. It registers in related word DB12c as a character string.
- the system control unit 14 of the related word registration server 10 selects the candidate character string ranked first. Not limited to, a predetermined number of candidate character strings such as the first to third ranking candidate character strings, the first and third ranking candidate character strings, or a predetermined ranking like the second ranking ranking A candidate character string may be stored.
- the related word registration server 10 functions as an example of related word registration means for registering the extracted registered character string and search word as related words in the related word database.
- step S19 If the difference is not greater than or equal to the threshold (step S19; NO), the related word registration server 10 ends the process.
- the information providing server 20 When the information providing server 20 performs a search based on the search query received from the user's terminal 30, as an example of the related word extraction unit, the information providing server 20 refers to the related word DB 12c of the related word registration server 10 and corresponds to the search word of the search query. Extract related terms. For example, the information providing server 20 transmits the received search query to the related word registration server 10. The related word registration server 10 receives the search query from the information providing server 20, and extracts a registered character string corresponding to the search word of the search query from the related word DB 12c. Then, the related word registration server 10 transmits the extracted registered character string as a related word to the information providing server 20 as an example of the related word output unit. The information providing server 20 transmits the received related word to the terminal 30.
- the terminal 30 displays the received related word in the related word display field 43 like “Maybe“ OO ”.
- a plurality of related words may be displayed, such as “related words of ranking 1” and “related words of ranking 2”. Moreover, not only when the number of search results is zero or small, related terms may be displayed.
- the related word registration server 10 calculates the distance between the extracted related word candidate character string and the search word (Ste S30). Specifically, the system control unit 14 of the related word registration server 10 calculates the distance Distance (w, u) between the search word u and the candidate character string w, for example, according to the equation (1).
- Distance (w, u) is a distance between the search word u and the candidate character string w, and is an index (an example of similarity) indicating how similar they are.
- Dist (w, u) is the Jaro-Winkler distance between the candidate character string w and the search word u
- Dist.Yomi (w, u) is the reading of the candidate character string w and the reading of the search word u.
- Distance (w, u) is preferably a Jaro-Winkler distance that can be expressed as a distance of 0 to 1. However, the distance is not limited to the Jaro-Winkler distance, and the similarity between the Levenshtein distance and the search word u and the candidate character string w. May be used as an index.
- the related word registration server 10 acquires the occurrence frequency of the candidate character string of each related word (step S31). Specifically, the system control unit 14 of the related word registration server 10 calculates the occurrence frequency of candidate character strings of each related word in the search query log. More specifically, the system control unit 14 of the related word registration server 10 calculates P (w) for each candidate character string w based on the data of the search query log DB 12a according to Expression (2).
- P (w) relates to the occurrence frequency of how much the candidate character string w is searched in the search log data of the search query log DB 12a.
- Occ.Count (w) is the number of appearances of the candidate character string w in the search query log (the number of times the candidate character string is used).
- a word having a high P (w) is a general word that is often input in searches by various users.
- the system control unit 14 of the related word registration server 10 calculates the suitability score based on the data of the search query storage unit (search query log DB 12a) as an example of the suitability score calculation unit.
- the related word registration server 10 calculates the searchability (availability) from the number of searches in the product DB 22a of the candidate character string of each related word (step S32). Specifically, first, the system control unit 14 of the related word registration server 10 transmits the candidate character string w of each related word to the information providing server 20, and the system control unit 24 of the information providing server 20 in the product DB 22a The number of searches (number of search results) is obtained, and the number of searches for each candidate character string w is transmitted to the related word registration server 10.
- the system control part 14 of the related word registration server 10 calculates the value of the search possibility of each candidate character string w from the number of search according to Formula (3).
- Availability (w) is a weighting factor for the number of search results (search hit count) for the search using the candidate character string w, and is expressed by, for example,
- #of Search Result (w) is the number of search results of the candidate character string w.
- FIG. 11 (A) the availability has various variations as shown in FIGS. 11 (B) and 11 (C) in addition to the formula (3).
- the related word registration server 10 determines each related word based on the distance between the search word and the related character candidate character string, the occurrence frequency of each related word candidate character string, and the search possibility of each related word candidate character string.
- the suitability score of the candidate character string is calculated (step S33).
- the system control unit 14 of the related word registration server 10 calculates the suitability score of each candidate character string w according to Expression (4).
- ⁇ and ⁇ are parameters for adjusting the suitability score.
- the parameter ⁇ is an adjustment value for P (w), and represents a degree of allowing correction to a word that is not so much input. Since the value of P (w) is small for words that are not input so much, it is difficult to pick up new words and new products by searching. If you want to be able to pick up words that are not yet used by searching as much as possible, increase the value of parameter ⁇ according to the server settings.
- the parameter ⁇ is an adjustment value for Distance (w, u) and mainly prevents Distance (w, u) from becoming zero and the score value to diverge.
- the related word registration server 10 determines the candidate character string based on the similarity between the candidate character string and the search word, the number of times the candidate character string is used, and the number of search results of the search using the candidate character string. It functions as an example of a score calculation means for calculating the suitability score. Further, the related word registration server 10 functions as an example of a parameter adjustment unit that adjusts information on the degree of similarity and the number of times of use.
- a related word candidate set obtained by extracting candidate character strings of related words based on one or more search results (an example of a predetermined condition) from the search query log DB 12a storing a log related to a search query.
- Generated and stored in the related word candidate DB 12b receives the search query of the search word input by the user, generates a partial character string from the character string of the search word, and related words in the related word candidate DB 12b based on the partial character string
- a candidate character string is extracted from the candidate set, and the suitability score of the candidate character string is based on the similarity between the candidate character string and the search word, the number of times the candidate character string is used, and the number of search results of the candidate character string.
- a reference line of the suitability score for the ranking is generated as a reference, and a candidate character string whose deviation between the suitability score and the reference line is equal to or greater than a preset threshold is extracted as a registered character string for registering as a related word and extracted. Since the candidate character string is extracted from the related word candidate set based on the partial character string having a reduced number of characters by registering the registered character string in the related word DB 12c as the related word of the search word, the suitability score of the candidate character string Thus, registration of unnecessary related words can be suppressed, and the accuracy of related words can be improved. Furthermore, according to the present embodiment, the processing can be speeded up.
- the search word when presented to the user based on the related word, the user can quickly reach the target search target.
- the related word registration server 10 sets the length of the partial character string based on the number of characters of the search word and generates the partial character string having the set length, the length of the partial character string is determined.
- the processing efficiency can be improved, the processing speed can be increased, and the precision of related words can be improved.
- the related word registration server 10 sets the movement width of the number of characters of the search word when generating the partial character string based on the number of characters of the search word, and generates the partial character string with the set movement width, Processing can be further accelerated.
- the related word registration server 10 when the related word registration server 10 generates a partial character string for a search word obtained by phonetic conversion of the search word, the partial character string can be efficiently extracted even with a short word, and the processing speed can be increased. The accuracy of related words can be improved.
- the related word registration server 10 further includes parameters ⁇ and ⁇ for adjusting the similarity of the score calculation unit and the information on the number of times of use as the parameter adjustment unit, the user can be adjusted by adjusting the parameters ⁇ and ⁇ .
- the suitability score can be changed according to the characteristics and circumstances of the, and the accuracy of related words can be further improved.
- the related word registration server 10 obtains a reference line from an approximate function that approximates the relation between the ranking and the suitability score as a determination reference generation unit, and as a result, an associated value registered character string as an abnormal value deviating from the approximate function. Can be easily detected, and the accuracy of related words can be further improved.
- the related word registration server 10 obtains the reference line L1 from the ranking data in a predetermined range (for example, higher rank) as the determination reference generation means, the candidate characters that are out of the reference line L1 from the higher ranking Since a column can be detected, a candidate character string in a higher ranking can be appropriately extracted.
- a predetermined range for example, higher rank
- the product DB 22a is further provided as an example of a search word storage means for storing a search word related to a product sold on an internal or external EC site, in particular, availability is easily made using data of the product DB 22a. Can be calculated and reflected in the suitability score.
- the related term registration server 10 further includes a search query log DB 12a as an example of a search query storage unit that stores a log related to a search query, and calculates a suitability score based on data in the search query log DB 12a
- the related word registration server 10 stores the search query log DB 12a. If candidate character strings are extracted from the words that have been created, it is possible to first narrow down to candidate character strings in which the user's interests are aggregated. In particular, based on the latest log, candidate character strings that particularly reflect user preferences can be extracted.
- the related word registration server 10 may set an upper limit on the number of registered character strings to be extracted as a registered character string extracting unit in step S19. In this case, registration of unnecessary related words can be suppressed, and the accuracy of related words can be improved.
- the information providing server 20 refers to the related word database and extracts and outputs related words corresponding to the search word of the received search query, the accuracy with which the user can quickly reach the target search target. High related words can be presented to the user.
- the reference line may be determined from the candidate character strings with the suitability score of the lower ranking among the extracted candidate character strings of the higher related words. More specifically, as shown in FIG. 9B, in the graph illustrating the relative suitability score for the ranking, the system control unit 14 of the related word registration server 10 has, for example, the suitability scores of the 50th place and the 10th place. Is determined (reference line L2).
- the relative suitability score is, for example, a value obtained by dividing each score by the value of the suitability score ranked first.
- the lower suitability score and the scores from the 50th to the 10th may be determined by the least square method.
- the lower suitability score is a suitability score that approximates each other in the ranking order as compared with the upper suitability score.
- the reference line L2 is an example of an approximation function that approximates the relationship between the ranking and the suitability score, and is a straight line connecting the suitability scores of the 50th place and the 10th place.
- the approximate function may be a logarithmic function, an exponential function, a power function such as a quadratic function or a cubic function, a function of a conic curve such as a hyperbola, etc. in addition to a straight line (linear function).
- the related word registration server 10 obtains the value of the intercept b with respect to the y-axis (relative adequacy score axis) of the reference line, and obtains a value obtained by multiplying the value of the intercept b by a constant.
- the threshold value ⁇ is assumed.
- the shopping site may be an external EC site connected to the network 3, and the information providing server 20 may be a general search server as well as a product search.
- the searchability is calculated from the number of search servers on the external EC site or the like.
- search query log DB12a stores a search query associated with the reception time of the search query together with the user identification information. Further, a search query is also stored in association with the category information of the product (category information will be described later).
- the user identification information for example, in the case of a search query from the terminal 30 of the user A who is logged in to the shopping site, the user ID of the user A or the header of the search query transmitted from the terminal 30 of the user A An existing IP address is listed.
- the reception time includes the reception time when the information providing server 20 receives the search query from the terminal 30, and the reception time when the related word registration server 10 receives the search query from the information providing server 20. If the information providing server 20 receives the search query from the terminal, the information providing server 20 also transmits the search query reception time to the related word registration server 10. Further, the reception time may be a time stamp added when the search query is stored in the search query log DB 12a.
- a user for example, user A logs in to a shopping site using the terminal 30 of user A, as shown in FIG. 6, information on the web page 40 that is the top page of the shopping site is obtained from the information providing server 20. It is transmitted to the terminal 30. Then, the web page 40 is displayed on the display unit 33 of the terminal 30.
- the Web page 40 is a top page of a shopping site, and a top category display 45 is displayed, and links to each category are provided. Furthermore, the category display 46 of the goods and services of the related group company of this shopping site is displayed, and the link to each category is made.
- the Web page 40 information "A's, Hello" indicating the user A is displayed.
- This search query is an example of the input search word, category information of the Web page 50 (category identification number indicating the category display 65 of the top category “home appliance”, etc.), and user identification information (user A) ID etc.
- the information providing server 20 receives this search query from the terminal 30, transmits it to the related word registration server 10, and searches for the product with reference to the product DB 22a. Further, the IP address of the transmission source of this search query is specified, and this IP address is also transmitted to the related word registration server 10.
- This IP address is an example of user identification information, and is used instead of the user ID when the user is not logged in.
- the related word registration server 10 receives a search query including a search word (step S41). Specifically, the system control unit 14 of the related word registration server 10 receives the search query from the information providing server 20 through the communication unit 11, and associates the received search query with the reception time of the search query in the search query log DB 12a.
- the system control unit 14 of the related word registration server 10 stores the search query in the search query log DB 12a in association with the IP address. Note that the system control unit 14 of the related word registration server 10 may include an IP address in the search query.
- the system control unit 14 of the related word registration server 10 may store the search query in the search query log DB 12a with the received order instead of the reception time. Good.
- the system control unit 14 of the related word registration server 10 functions as an example of a search query storage unit that stores the received search queries in accordance with the reception order.
- the related word registration server 10 determines whether or not a search query having a matching user ID exists in the search query log (step S42). Specifically, the system control unit 14 of the related word registration server 10 searches for a search query that coincides with the user ID of the user A and whose reception time is earlier than the received search query (an example of reception order earlier). The query log DB 12a is extracted. Then, the system control unit 14 of the related word registration server 10 determines that the previous search query exists in the search query log when the previous search query can be extracted, and determines that the previous search query does not exist when it cannot be extracted. To do. Note that the system control unit 14 of the related word registration server 10 may extract the search query log DB 12a for a search query that matches the IP address of the received search query and whose reception time is earlier than the received search query.
- the case of matching with the user ID or the case of matching with the IP address is an example of preset search query extraction conditions.
- the related term registration server 10 functions as an example of a search query extraction unit that extracts a search query that matches the user identification information of the user who has input the search query as the search query extraction condition.
- the related word registration server 10 determines whether the reception time and the reception time of the previous search query are within a predetermined time. (Step S43). Specifically, the system control unit 14 of the related word registration server 10 determines that the difference between the reception time of the received search query and the reception time of the previous search query is within a predetermined time (for example, 3 to 60 seconds). The previous search query in between) is extracted. Then, the system control unit 14 of the related word registration server 10 determines that the previous search query is within a predetermined time when the previous search query can be extracted, and determines that the previous search query is not within the predetermined time when it cannot be extracted.
- a predetermined time for example, 3 to 60 seconds
- the predetermined time is an example of a preset search query extraction condition.
- the related term registration server 10 is based on a search query extraction condition that is set in advance, and the search query extraction unit that extracts a search query having a reception time earlier than the received search query from the search query storage unit. It serves as an example. Further, the related term registration server 10 functions as an example of a search query extraction unit that extracts a previous search query whose reception time is within a predetermined time from the reception time of the search query as a search query extraction condition.
- the related word registration server 10 determines that the received search word is the closest search word with the closest reception time.
- the character string set is stored (step S44). Specifically, the system control unit 14 of the related word registration server 10 configures a search query having the closest reception time to a search word constituting the received search query among the extracted search queries. A search word is selected, and a character string set of the previous search word and the received search word is stored in the character string set DB 12d.
- system control unit 14 of the related word registration server 10 may further narrow down to a search query that matches the category information (for example, the top category) as an example of a search query extraction condition set in advance.
- the related word registration server 10 stores the search query in the search query log in association with the user ID and the reception time (step S45). Specifically, the system control unit 14 of the related word registration server 10 stores the search query in the search query log DB 12a in association with the category information, the user ID (may be an IP address) and the reception time. In addition, when a search query having a matching user ID does not exist in the search query log (step S42; NO), or when the reception time and the reception time of the previous search query are not within a predetermined time (step S43; NO). In addition, the system control unit 14 of the related word registration server 10 stores the search query in the search query log DB 12a in association with the user ID and the reception time.
- the related word registration server 10 is an example of a character string set storage unit that stores a search word that constitutes the extracted search query and a search word that constitutes the received search query as a character string set. Function as.
- the information on the Web page 60 is transmitted to the terminal 30 as a search result as shown in FIG.
- the terminal 30 receives the information of the Web page 60 from the information providing server 20 and displays it on the display unit 33.
- the user A looks at the search result of the Web page 60, and if there is a target product, clicks on the product in the product display column 63. If there is no target product, the user A further inputs a search word into the search word input column 61.
- the terminal 30 transmits a search query including the search word “ ⁇ O ⁇ ” and category information (“home appliance”) to the information providing server 20 as described above.
- the information provision server 20 receives this search query, transmits to the related word registration server 10, and performs a search.
- the related word registration server 10 receives the search query as described above (step S41), and performs the processing from step S42 to step S45.
- the above-described processing is also performed for a search query from another user B or the like, and a search query log of the search query associated with the reception time together with the category information, each user ID, and each IP address is generated in the search query log DB 12a. Then, a character string set set in which the character strings are collected is generated in the character string set DB 12d.
- the related word registration server 10 determines whether or not the extraction start condition is satisfied (step S50). Specifically, the system control unit 14 of the related word registration server 10 stores, as an example of the extraction start condition, the number of search queries (total number of search queries) of the search query log in the search query log DB 12a and the character string set DB 12d. It is determined whether or not the number of elements in the character string set set (total number of character string set sets) exceeds a predetermined threshold. Further, the system control unit 14 of the related word registration server 10 may determine whether or not a predetermined time has elapsed since the character string set extraction start condition was satisfied last time.
- the related word registration server 10 extracts a character string set having the same search word from the character string set set (step S51). Specifically, as shown in FIG. 16, the system control unit 14 of the related word registration server 10 extracts a character string set 70 in which the previous search word 71 is the same (the search word is the same) from the character string set DB 12d. To do. If the extraction start condition is not satisfied (step S50; NO), the related word registration process is terminated.
- the system control unit 14 of the related word registration server 10 may extract a character string set having a similar previous search word from the character set DB 12d according to a preset character string set extraction start condition.
- the search words are similar if, for example, the similarity value based on the distance between character strings such as the Jaro-Winker distance and the Levenstein distance is equal to or greater than a predetermined threshold.
- the system control unit 14 of the related word registration server 10 also extracts a character string set having the previous search word “ ⁇ ⁇ ” that is similar to the previous search word 71.
- the related word registration server 10 extracts a character string set whose character string set is the same or similar to the previous search word from the character set set according to a preset character string set extraction start condition. Functions as an example.
- the related word registration server 10 determines whether or not the total number of extracted character string sets is equal to or greater than a predetermined value (step S52). Specifically, the system control unit 14 of the related word registration server 10 determines whether or not the total number of character string groups related to a certain search word 71 is greater than or equal to a predetermined value (for example, 3 or more). As shown in FIG. 16, the character string set relating to the previous search word 71 is the character string set 70 of ⁇ previous search word 71-search word 71>, and the character string set of ⁇ previous search word 71-search word 72>. 70, and a total of three character string sets 70 of ⁇ previous search word 71-search word 73>.
- a predetermined value for example, 3 or more
- the related word registration server 10 determines whether there is a character string set that satisfies the registration condition among the extracted character string sets (step S52). S53). Specifically, the system control unit 14 of the related word registration server 10 determines whether or not the number of each character string set 70 is a predetermined number or more (for example, 10 or more) as an example of the registration condition. As described above, the system control unit 14 of the related word registration server 10 sets the character string set having the same or similar search words 72, 73, 74 in the character string set 70 having the same or similar search word 71. It is determined whether the number has exceeded a predetermined threshold.
- the system control unit 14 of the related word registration server 10 specifies a character string set that exceeds a predetermined threshold as a related word. It should be noted that the system control unit 14 of the related word registration server 10 may count the same character string set if there is a search word “ ⁇ ⁇ ” similar to the search word 73 “ ⁇ ”.
- the related word registration server 10 functions as an example of a related word specifying unit that specifies a character string set as a related word from the extracted character string set based on preset registration conditions.
- the related word registration server 10 has, as a registration condition, a case where the number of character string sets having the same or similar search word exceeds a predetermined threshold in a character string set having the same or similar previous search word.
- it functions as an example of related word specifying means for specifying the character string set as a related word.
- the system control unit 14 of the related word registration server 10 has a ratio of character string sets having the same or similar search word in character string sets having the same or similar search word 71. It may be determined whether or not a predetermined threshold value (for example, 80%) is exceeded.
- a predetermined threshold value for example, 80%
- the related word registration server 10 has, as a registration condition, the ratio of character string sets having the same or similar search word in a character string set having the same or similar previous search word exceeds a predetermined threshold.
- a predetermined threshold for specifying the character string set as a related word.
- the related word registration server 10 registers the character string set as a related word (step S54).
- the system control unit 14 of the related word registration server 10 includes three (15%) character string groups 70 of ⁇ previous search word 71-search word 71>.
- the character string set 70 of ⁇ previous search word 71-search word 72> is registered in the related word DB 12c as a related word.
- the related word registration server 10 functions as an example of related word registration means for registering the specified character string set as a related word.
- the related word registration server 10 determines whether or not a character string set that does not satisfy the registration condition is registered in the related word DB (step S53). S55). Specifically, the system control unit 14 of the related word registration server 10 reduces the ratio of the previously registered ⁇ previous search word-search word> character string set 70 in the character string set DB 12d. It is determined whether or not a predetermined threshold value (for example, 80%) has been cut. This occurs when the number of other character string sets having the same or similar previous search word increases.
- a predetermined threshold value for example, 80%
- the related word registration server 10 deletes the character string set that does not satisfy the registration condition from the related word DB (step S56). Specifically, the system control unit 14 of the related word registration server 10 deletes, from the related word DB 12c, a character string set that does not satisfy the registration condition from the already registered character string sets.
- the related word registration server 10 determines whether or not all character string groups have been extracted (step S57). Specifically, the system control unit 14 of the related word registration server 10 determines whether or not the character string set extraction processing in step S51 has been performed on all previous search words in the character string set set of the character string set DB 12d. If not extracted for all character string groups (step S57; NO), the process returns to step S11, and if extracted for all character string groups (step S57; YES), related words are registered. Terminate the process.
- the search query of the search word input by the user is received, and the previous search query acquired earlier in time than the received search query is based on the preset search query extraction condition.
- the search word extracted from the search query log, the search word of the previous search query and the search word of the received search query are stored as a character string set in the character string set DB 12d, and a preset character string set extraction start condition Is received by extracting the character string set having the same or similar previous search word and registering the extracted character string set satisfying the preset registration condition as the related word in the related word DB 12c.
- a certain amount of character string pairs of the search query and the previous search query based on the search query extraction conditions is accumulated as the character string combination extraction start condition, and the stored character string pairs are registered.
- the related word registration server 10 further stores user identification information and extracts a search query that matches the user identification information of the user who has input the search query as a search query extraction condition, a character string for the same user Since a set can be generated, the accuracy of related words can be further improved.
- the related word registration server 10 extracts a previous search query whose reception time is within a predetermined time from the reception time of the search query as a search query extraction condition, Since the search query can be extracted, the accuracy of related terms can be further improved.
- the related word registration server 10 further stores category information and extracts a previous search query based on the category information as a search query extraction condition, search words having the same or related category are grouped into character strings. Therefore, the accuracy of related words can be further improved.
- the related word registration server 10 extracts a character string set when the number of elements of the character string set exceeds a predetermined threshold as a character string set extraction start condition, a character string set of a certain scale is obtained. Since it can be stored and an appropriate character string set can be extracted from it, the accuracy of related words can be further improved.
- the related word registration server 10 has a registration condition, the number of character string sets having the same or similar search word in a character string set having the same or similar previous search word exceeds a predetermined threshold If a character string set is registered as a related word, a highly accurate character string set can be selected, and the accuracy of the related word can be further improved.
- the related word registration server 10 has, as a registration condition, the ratio of character string sets having the same or similar search word in a character string set having the same or similar previous search word exceeds a predetermined threshold If the character string set is registered as a related word, a highly accurate character string set can be selected, and the accuracy of the related word can be further improved.
- the related word registration server 10 may extract a character string set in step S51 when the number of search queries in the search query log DB 12a exceeds a predetermined threshold as a character string set extraction start condition.
- a character string set of a certain scale can be accumulated, and an appropriate character string set can be extracted from the character string set. Therefore, the precision of related words can be further improved.
- the related word registration server 10 may extract a character string set when a predetermined time has elapsed after satisfying the character string set extraction start condition in step S50.
- a character string set of a certain scale can be accumulated, and an appropriate character string set can be extracted from the character string set. Therefore, the precision of related words can be further improved.
- the related words may be registered by combining the operations of the first and second embodiments in the related word registration system.
- a related word in which the distance between character strings is considered as in the operation of the first embodiment and a related word in which the order in which the search query is used is considered as in the operation in the second embodiment.
- the related word database may be constructed in the related word DB 12c.
- the information providing server 20 is constructed in the related word DB 12 c by the operation of the first embodiment.
- the related word database as shown in FIG. 6, “probably“ XXX ”is displayed on the Web page 40 as the related word.
- the information providing server 20 refers to the related word database constructed by the operation of the second embodiment in the related word DB 12c. Then, as a related word, “? ...?” May be displayed on the Web page 40.
- these related words may be displayed simultaneously.
- the present invention is not limited to the above embodiments.
- Each of the embodiments described above is an exemplification, and any configuration that has substantially the same configuration as the technical idea described in the claims of the present invention and has the same operational effects can be used. It is included in the technical scope of the present invention.
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Mathematical Physics (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Audiology, Speech & Language Pathology (AREA)
- General Health & Medical Sciences (AREA)
- Computer Hardware Design (AREA)
- Software Systems (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
Description
(2.1 関連語登録サーバ10の構成および機能)
次に、関連語登録サーバ10の構成および機能について、図2を用いて説明する。
次に、情報提供サーバ20の構成および機能について、図3を用いて説明する。
図3は、情報提供サーバ20の概要構成の一例を示すブロック図である。
次に、端末30の構成および機能について、図4を用いて説明する。
図4は、端末30の概要構成の一例を示すブロック図である。
次に、本発明の一実施形態に係る関連語登録システム1における第1実施形態の動作について図5から図11を用い説明する。
次に、関連語の登録の動作について、図6~図9を用いて説明する。
次に、適否スコアの算出のサブルーチンについて、図10および図11を用いて説明する。
次に、本発明の一実施形態に係る関連語登録システム1における第2実施形態の動作について、図6、および、図12から図16を用い説明する。なお、前記第1実施形態の動作と同一または対応する部分には、同一の符号を用いて異なるところを主に説明する。その他の実施形態および変形例も同様とする。
まず、検索クエリログおよび文字列組集合の生成について、図6、および、図12から図14を用いて説明する。
次に、関連語の登録について、図15および図16を用いて説明する。
10:関連語登録サーバ(関連語登録装置)
12a:検索クエリログDB
12b:関連語候補DB
12c:関連語DB
12d:文字列組DB
20:情報提供サーバ(情報処理装置)
22a:商品DB
30:端末
Claims (20)
- 検索クエリに関するログを記憶した検索クエリログから、所定の条件に基づいて関連語の候補を抽出して関連語候補集合を生成する関連語候補集合生成手段と、
ユーザが入力した検索ワードの検索クエリを受信する受信手段と、
前記検索ワードの文字列から部分文字列を生成する部分文字列生成手段と、
前記生成された部分文字列に基づいて、前記関連語候補集合から候補文字列を抽出する候補文字列抽出手段と、
前記候補文字列と前記検索ワードとの類似度、前記候補文字列の使用回数、および、前記候補文字列による検索の検索結果数とに基づいて、前記候補文字列の適否スコアを算出するスコア算出手段と、
前記スコアの順に前記候補文字列に対するランキングを行うランキング手段と、
前記候補文字列の適否スコアとランキングとに基づき、前記候補文字列を関連語として登録するか否かの判定基準として前記ランキングに対する適否スコアの基準ラインを生成する判定基準生成手段と、
前記適否スコアと基準ラインとの乖離が予め設定された閾値以上である候補文字列を関連語として登録するための登録文字列として抽出する登録文字列抽出手段と、
前記抽出された登録文字列と前記検索ワードとを、関連語データベースに関連語として登録する関連語登録手段と、
を備えた関連語登録装置。 - 請求項1に記載の関連語登録装置において、
前記部分文字列生成手段が、前記検索ワードの文字数に基づいて部分文字列の長さを設定し、当該設定された長さの部分文字列を生成することを特徴とする関連語登録装置。 - 請求項1に記載の関連語登録装置において、
前記部分文字列生成手段が、前記検索ワードの文字数に基づいて部分文字列を生成する際の検索ワードの文字数の移動幅を設定し、当該設定された移動幅で部分文字列を生成することを特徴とする関連語登録装置。 - 請求項1から請求項3のいずれか1項に記載の関連語登録装置において、
前記部分文字列生成手段が、前記検索ワードを表音変換させた検索ワードに対して、前記部分文字列を生成することを特徴とする関連語登録装置。 - 請求項1から請求項4のいずれか1項に記載の関連語登録装置において、
前記スコア算出手段の前記類似度と前記使用回数の情報とを調節するパラメータ調節手段を更に備えたことを特徴とする関連語登録装置。 - 請求項1から請求項5のいずれか1項に記載の関連語登録装置において、
前記判定基準生成手段が、前記ランキングと前記適否スコアとの関連を近似する近似関数より、または、所定の範囲の前記ランキングのデータより、前記基準ラインを求めることを特徴とする関連語登録装置。 - 請求項1から請求項6のいずれか1項に記載の関連語登録装置において、
前記登録文字列抽出手段が、抽出する前記登録文字列の数に上限を設けることを特徴とする関連語登録装置。 - 請求項1から請求項7のいずれか1項に記載の関連語登録装置において、
内部又は外部のECサイトにおいて販売されている商品に関連した検索ワードを記憶する検索ワード記憶手段を更に備えたことを特徴とする関連語登録装置。 - 請求項1から請求項8のいずれか1項に記載の関連語登録装置において、前記スコア算出手段が、前記検索クエリに関するログのデータに基づき、前記適否スコアを算出すること特徴とする関連語登録装置。
- 請求項1から請求項9のいずれか1項に記載の関連語登録装置において、
前記受信した検索クエリを、受信順に従って記憶する検索クエリ記憶手段と、
予め設定した検索クエリ抽出条件に基づいて、前記受信した検索クエリより受信順が早い先の検索クエリを、前記検索クエリ記憶手段から抽出する検索クエリ抽出手段と、
前記抽出した先の検索クエリを構成する先の検索ワードと、前記受信した検索クエリを構成する検索ワードと、を文字列組として記憶する文字列組記憶手段と、
予め設定した文字列組抽出開始条件に応じて、前記文字列組記憶手段から前記先の検索ワードが同一または類似である文字列組を抽出する文字列抽出手段と、
予め設定した登録条件に基づいて、前記抽出した文字列組から関連語となる文字列組を特定する関連語特定手段と、
を更に備え、
前記関連語登録手段が、前記特定された文字列組を関連語として関連語データベースに登録することを特徴とする関連語登録装置。 - 請求項10に記載の関連語登録装置において、
前記検索クエリ記憶手段は、ユーザ識別情報をさらに記憶し、
前記検索クエリ抽出手段が、前記検索クエリ抽出条件として、前記検索クエリを入力したユーザのユーザ識別情報が一致する前記先の検索クエリを抽出することを特徴とする関連語登録装置。 - 請求項10または請求項11に記載の関連語登録装置において、
前記検索クエリ抽出手段が、前記検索クエリ抽出条件として、前記検索クエリの受信時刻から所定の時間内に受信した先の検索クエリを抽出することを特徴とする関連語登録装置。 - 請求項10から請求項12のいずれか1項に記載の関連語登録装置において、
前記検索クエリ記憶手段は、カテゴリ情報をさらに記憶し、
前記検索クエリ抽出手段が、前記検索クエリ抽出条件として、前記カテゴリ情報に基づき、前記先の検索クエリを抽出することを特徴とする関連語登録装置。 - 請求項10から請求項13いずれか1項に記載の関連語登録装置において、
前記文字列抽出手段が、前記文字列組抽出開始条件として、前記検索クエリ記憶手段における検索クエリ数が、または、前記文字列組集合の要素数が、所定の閾値を超えた場合に、前記文字列組を抽出することを特徴とする関連語登録装置。 - 請求項10から請求項13のいずれか1項に記載の関連語登録装置において、
前記文字列抽出手段が、前記文字列組抽出開始条件を満たしてから所定の時間を経過した場合に、前記文字列組を抽出することを特徴とする関連語登録装置。 - 請求項10から請求項15のいずれか1項に記載の関連語登録装置において、
前記関連語特定手段が、前記登録条件として、前記先の検索ワードが同一または類似である文字列組において、同一または類似の検索ワードを有する文字列組の数、または、同一または類似の検索ワードを有する文字列組の割合が、所定の閾値を超えた場合に、当該文字列組を関連語として特定することを特徴とする関連語登録装置。 - 請求項1から請求項16のいずれか1項に記載の関連語登録装置の関連語データベースを参照して情報処理する情報処理装置であって、
前記関連語データベースを参照して、前記受信した検索クエリの検索ワードに対応する関連語を抽出する関連語抽出手段と、
前記関連語抽出手段により抽出された関連語を出力する関連語出力手段と、
を備えたことを特徴とする情報処理装置。 - 関連語を登録する関連語登録装置の関連語登録方法において、
検索クエリに関するログを記憶した検索クエリログから、所定の条件に基づいて関連語の候補を抽出して関連語候補集合を生成する関連語候補集合生成ステップと、
ユーザが入力した検索ワードの検索クエリを受信する受信ステップと、
前記検索ワードの文字列から部分文字列を生成する部分文字列生成ステップと、
前記生成された部分文字列に基づいて、前記関連語候補集合から候補文字列を抽出する候補文字列抽出ステップと、
前記候補文字列と前記検索ワードとの類似度、前記候補文字列の使用回数、および、前記候補文字列による検索の検索結果数とに基づいて、前記候補文字列の適否スコアを算出するスコア算出ステップと、
前記スコアの順に前記候補文字列に対するランキングを行うランキングステップと、
前記候補文字列の適否スコアとランキングとに基づき、前記候補文字列を関連語として登録するか否かの判定基準として前記ランキングに対する適否スコアの基準ラインを生成する判定基準生成ステップと、
前記適否スコアと基準ラインとの乖離が予め設定された閾値以上である候補文字列を関連語として登録するための登録文字列として抽出する登録文字列抽出ステップと、
前記抽出された登録文字列と前記検索ワードとを、関連語データベースに関連語として登録する関連語登録ステップと、
を含む関連語登録方法。 - コンピュータを、
検索クエリに関するログを記憶した検索クエリログから、所定の条件に基づいて関連語の候補を抽出して関連語候補集合を生成する関連語候補集合生成手段、
ユーザが入力した検索ワードの検索クエリを受信する受信手段、
前記検索ワードの文字列から部分文字列を生成する部分文字列生成手段、
前記生成された部分文字列に基づいて、前記関連語候補集合から候補文字列を抽出する候補文字列抽出手段、
前記候補文字列と前記検索ワードとの類似度、前記候補文字列の使用回数、および、前記候補文字列による検索の検索結果数とに基づいて、前記候補文字列の適否スコアを算出するスコア算出手段、
前記スコアの順に前記候補文字列に対するランキングを行うランキング手段、
前記候補文字列の適否スコアとランキングとに基づき、前記候補文字列を関連語として登録するか否かの判定基準として前記ランキングに対する適否スコアの基準ラインを生成する判定基準生成手段、
前記適否スコアと基準ラインとの乖離が予め設定された閾値以上である候補文字列を関連語として登録するための登録文字列として抽出する登録文字列抽出手段、および、
前記抽出された登録文字列と前記検索ワードとを関連語として関連語データベースに登録する関連語登録手段として機能させる関連語登録装置用プログラム。 - コンピュータを、
検索クエリに関するログを記憶した検索クエリログから、所定の条件に基づいて関連語の候補を抽出して関連語候補集合を生成する関連語候補集合生成手段、
ユーザが入力した検索ワードの検索クエリを受信する受信手段、
前記検索ワードの文字列から部分文字列を生成する部分文字列生成手段、
前記生成された部分文字列に基づいて、前記関連語候補集合から候補文字列を抽出する候補文字列抽出手段、
前記候補文字列と前記検索ワードとの類似度、前記候補文字列の使用回数、および、前記候補文字列による検索の検索結果数とに基づいて、前記候補文字列の適否スコアを算出するスコア算出手段、
前記スコアの順に前記候補文字列に対するランキングを行うランキング手段、
前記候補文字列の適否スコアとランキングとに基づき、前記候補文字列を関連語として登録するか否かの判定基準として前記ランキングに対する適否スコアの基準ラインを生成する判定基準生成手段、
前記適否スコアと基準ラインとの乖離が予め設定された閾値以上である候補文字列を関連語として登録するための登録文字列として抽出する登録文字列抽出手段、および、
前記抽出された登録文字列と前記検索ワードとを、関連語データベースに関連語として登録する関連語登録手段として機能させる関連語登録装置用の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。
Priority Applications (11)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
KR1020137015030A KR101361403B1 (ko) | 2010-11-10 | 2011-11-07 | 관련어 등록 장치, 정보 처리 장치, 관련어 등록 방법 및 기록 매체 |
ES11839828.8T ES2577938T3 (es) | 2010-11-10 | 2011-11-07 | Dispositivo de registro de palabras relacionadas, dispositivo de procesamiento de información, método de registro de palabras relacionadas, programa para dispositivo de registro de palabras relacionadas, y medio de almacenamiento |
BR122013013420-4A BR122013013420B1 (pt) | 2010-11-10 | 2011-11-07 | dispositivo e método de registro de palavra relacionada |
KR1020137014718A KR101368594B1 (ko) | 2010-11-10 | 2011-11-07 | 관련어 등록 장치, 정보 처리 장치, 관련어 등록 방법 및 기록 매체 |
CN201180053722.8A CN103201737B (zh) | 2010-11-10 | 2011-11-07 | 关联词登记装置、信息处理装置、关联词登记方法、关联词登记装置用程序以及记录介质 |
US13/884,344 US8606565B2 (en) | 2010-11-10 | 2011-11-07 | Related-word registration device, information processing device, related-word registration method, program for related-word registration device, and recording medium |
CA2817131A CA2817131C (en) | 2010-11-10 | 2011-11-07 | Related-word registration device, information processing device, related-word registration method, program for related-word registration device, and recording medium |
JP2012533417A JP5101759B2 (ja) | 2010-11-10 | 2011-11-07 | 関連語登録装置、情報処理装置、関連語登録方法、関連語登録装置用プログラム、および、記録媒体 |
EP11839828.8A EP2639705B1 (en) | 2010-11-10 | 2011-11-07 | Related-word registration device, information processing device, related-word registration method, program for related-word registration device, and recording medium |
BR112013011573-4A BR112013011573B1 (pt) | 2010-11-10 | 2011-11-07 | dispositivo e método de registro de palavra relacionada, sistema servidor e mídia de gravação legível por computador |
US13/923,744 US8738366B2 (en) | 2010-11-10 | 2013-06-21 | Related-word registration device, information processing device, related-word registration method, program for related-word registration device, and recording medium |
Applications Claiming Priority (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2010252325 | 2010-11-10 | ||
JP2010252326 | 2010-11-10 | ||
JP2010-252326 | 2010-11-10 | ||
JP2010-252325 | 2010-11-10 |
Related Child Applications (2)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US13/884,344 A-371-Of-International US8606565B2 (en) | 2010-11-10 | 2011-11-07 | Related-word registration device, information processing device, related-word registration method, program for related-word registration device, and recording medium |
US13/923,744 Division US8738366B2 (en) | 2010-11-10 | 2013-06-21 | Related-word registration device, information processing device, related-word registration method, program for related-word registration device, and recording medium |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2012063770A1 true WO2012063770A1 (ja) | 2012-05-18 |
Family
ID=46050916
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2011/075572 WO2012063770A1 (ja) | 2010-11-10 | 2011-11-07 | 関連語登録装置、情報処理装置、関連語登録方法、関連語登録装置用プログラム、および、記録媒体 |
Country Status (9)
Country | Link |
---|---|
US (2) | US8606565B2 (ja) |
EP (2) | EP2650805B1 (ja) |
JP (2) | JP5101759B2 (ja) |
KR (2) | KR101368594B1 (ja) |
CN (2) | CN103279557B (ja) |
BR (2) | BR122013013420B1 (ja) |
CA (2) | CA2822273C (ja) |
ES (2) | ES2577938T3 (ja) |
WO (1) | WO2012063770A1 (ja) |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2014106832A (ja) * | 2012-11-29 | 2014-06-09 | Nec Software Tohoku Ltd | 情報提供装置及び情報提供方法 |
CN106407764A (zh) * | 2016-09-30 | 2017-02-15 | 深圳天珑无线科技有限公司 | 信息生成方法及装置 |
Families Citing this family (24)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US10417646B2 (en) | 2010-03-09 | 2019-09-17 | Sdl Inc. | Predicting the cost associated with translating textual content |
US8606565B2 (en) | 2010-11-10 | 2013-12-10 | Rakuten, Inc. | Related-word registration device, information processing device, related-word registration method, program for related-word registration device, and recording medium |
US11003838B2 (en) | 2011-04-18 | 2021-05-11 | Sdl Inc. | Systems and methods for monitoring post translation editing |
US10261994B2 (en) * | 2012-05-25 | 2019-04-16 | Sdl Inc. | Method and system for automatic management of reputation of translators |
JP6068901B2 (ja) * | 2012-09-26 | 2017-01-25 | 京セラ株式会社 | 情報端末、音声操作プログラムおよび音声操作方法 |
US9280536B2 (en) * | 2013-03-28 | 2016-03-08 | Hewlett Packard Enterprise Development Lp | Synonym determination among n-grams |
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 |
GB2549240A (en) * | 2015-01-06 | 2017-10-18 | What3Words Ltd | A method for suggesting one or more multi-word candidates based on an input string received at an electronic device |
GB2535439A (en) * | 2015-01-06 | 2016-08-24 | What3Words Ltd | A method for suggesting candidate words as replacements for an input string received at an electronic device |
JP6621652B2 (ja) * | 2015-11-18 | 2019-12-18 | シャープ株式会社 | 電子機器、およびその制御方法 |
WO2018009369A1 (en) * | 2016-07-06 | 2018-01-11 | Mastercard International Incorporated | Method and system for providing sales information and insights through a conversational interface |
CN106528534A (zh) * | 2016-11-09 | 2017-03-22 | 天津赛因哲信息技术有限公司 | 基于专有名词的关联词提取方法 |
US10719539B2 (en) * | 2017-06-06 | 2020-07-21 | Mastercard International Incorporated | Method and system for automatic reporting of analytics and distribution of advice using a conversational interface |
US11537644B2 (en) * | 2017-06-06 | 2022-12-27 | Mastercard International Incorporated | Method and system for conversational input device with intelligent crowd-sourced options |
JP6762678B2 (ja) | 2018-03-27 | 2020-09-30 | 日本電信電話株式会社 | 違法コンテンツ探索装置、違法コンテンツ探索方法およびプログラム |
JP6947307B2 (ja) * | 2018-07-25 | 2021-10-13 | 日本電信電話株式会社 | 解析装置、解析方法及び解析プログラム |
JP7443667B2 (ja) * | 2019-03-25 | 2024-03-06 | カシオ計算機株式会社 | 検索装置、辞書検索プログラム、辞書検索方法 |
JP6780129B1 (ja) * | 2019-03-27 | 2020-11-04 | 三菱電機ビルテクノサービス株式会社 | 設備機器情報収集システム |
JP7256935B2 (ja) * | 2019-09-02 | 2023-04-13 | 富士通株式会社 | 辞書作成装置及び辞書作成方法 |
CN111261165B (zh) * | 2020-01-13 | 2023-05-16 | 佳都科技集团股份有限公司 | 车站名称识别方法、装置、设备及存储介质 |
US11776529B2 (en) * | 2020-04-28 | 2023-10-03 | Samsung Electronics Co., Ltd. | Method and apparatus with speech processing |
KR20210132855A (ko) * | 2020-04-28 | 2021-11-05 | 삼성전자주식회사 | 음성 처리 방법 및 장치 |
JP7426302B2 (ja) * | 2020-06-30 | 2024-02-01 | 日立建機株式会社 | 同義語生成装置、及び同義語生成プログラム |
CN112818262B (zh) * | 2021-01-28 | 2023-07-21 | 上海博泰悦臻网络技术服务有限公司 | 基于用户数据的地图poi搜索方法、系统、设备及介质 |
Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2000268035A (ja) | 1999-03-12 | 2000-09-29 | Nippon Telegr & Teleph Corp <Ntt> | 意味辞書登録装置 |
JP2010231433A (ja) * | 2009-03-26 | 2010-10-14 | Fujitsu Ten Ltd | 検索装置 |
Family Cites Families (19)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6457004B1 (en) * | 1997-07-03 | 2002-09-24 | Hitachi, Ltd. | Document retrieval assisting method, system and service using closely displayed areas for titles and topics |
JP3219386B2 (ja) * | 1997-12-26 | 2001-10-15 | 松下電器産業株式会社 | 情報フィルタ装置及び情報フィルタ方法 |
US6006225A (en) * | 1998-06-15 | 1999-12-21 | Amazon.Com | Refining search queries by the suggestion of correlated terms from prior searches |
JP3696745B2 (ja) * | 1999-02-09 | 2005-09-21 | 株式会社日立製作所 | 文書検索方法及び文書検索システム及び文書検索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 |
US6314419B1 (en) * | 1999-06-04 | 2001-11-06 | Oracle Corporation | Methods and apparatus for generating query feedback based on co-occurrence patterns |
DE19952769B4 (de) * | 1999-11-02 | 2008-07-17 | Sap Ag | Suchmaschine und Verfahren zum Abrufen von Informationen mit Abfragen in natürlicher Sprache |
JP4342753B2 (ja) * | 2001-08-10 | 2009-10-14 | 株式会社リコー | 文書検索装置、文書検索方法、プログラム及びコンピュータに読み取り可能な記憶媒体 |
US7308404B2 (en) * | 2001-09-28 | 2007-12-11 | Sri International | Method and apparatus for speech recognition using a dynamic vocabulary |
JP4226862B2 (ja) * | 2002-08-29 | 2009-02-18 | 株式会社リコー | 文書検索装置 |
US7885963B2 (en) * | 2003-03-24 | 2011-02-08 | Microsoft Corporation | Free text and attribute searching of electronic program guide (EPG) data |
US7454393B2 (en) * | 2003-08-06 | 2008-11-18 | Microsoft Corporation | Cost-benefit approach to automatically composing answers to questions by extracting information from large unstructured corpora |
US7636714B1 (en) * | 2005-03-31 | 2009-12-22 | Google Inc. | Determining query term synonyms within query context |
US8438142B2 (en) * | 2005-05-04 | 2013-05-07 | Google Inc. | Suggesting and refining user input based on original user input |
US7657518B2 (en) * | 2006-01-31 | 2010-02-02 | Northwestern University | Chaining context-sensitive search results |
JP2008250625A (ja) * | 2007-03-30 | 2008-10-16 | Nomura Research Institute Ltd | 検索システム |
CN100476800C (zh) * | 2007-06-22 | 2009-04-08 | 腾讯科技(深圳)有限公司 | 一种切分索引分词的方法及系统 |
CN101206673A (zh) * | 2007-12-25 | 2008-06-25 | 北京科文书业信息技术有限公司 | 网络搜索过程中关键词的智能纠错系统及方法 |
US8346534B2 (en) * | 2008-11-06 | 2013-01-01 | University of North Texas System | Method, system and apparatus for automatic keyword extraction |
US8606565B2 (en) | 2010-11-10 | 2013-12-10 | Rakuten, Inc. | Related-word registration device, information processing device, related-word registration method, program for related-word registration device, and recording medium |
-
2011
- 2011-11-07 US US13/884,344 patent/US8606565B2/en active Active
- 2011-11-07 WO PCT/JP2011/075572 patent/WO2012063770A1/ja active Application Filing
- 2011-11-07 EP EP13175515.9A patent/EP2650805B1/en active Active
- 2011-11-07 BR BR122013013420-4A patent/BR122013013420B1/pt active IP Right Grant
- 2011-11-07 CN CN201310231780.6A patent/CN103279557B/zh active Active
- 2011-11-07 KR KR1020137014718A patent/KR101368594B1/ko active IP Right Grant
- 2011-11-07 CA CA2822273A patent/CA2822273C/en active Active
- 2011-11-07 ES ES11839828.8T patent/ES2577938T3/es active Active
- 2011-11-07 CA CA2817131A patent/CA2817131C/en active Active
- 2011-11-07 CN CN201180053722.8A patent/CN103201737B/zh active Active
- 2011-11-07 EP EP11839828.8A patent/EP2639705B1/en active Active
- 2011-11-07 JP JP2012533417A patent/JP5101759B2/ja active Active
- 2011-11-07 KR KR1020137015030A patent/KR101361403B1/ko active IP Right Grant
- 2011-11-07 BR BR112013011573-4A patent/BR112013011573B1/pt active IP Right Grant
- 2011-11-07 ES ES13175515.9T patent/ES2642379T3/es active Active
-
2012
- 2012-07-24 JP JP2012163636A patent/JP5170804B2/ja active Active
-
2013
- 2013-06-21 US US13/923,744 patent/US8738366B2/en active Active
Patent Citations (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2000268035A (ja) | 1999-03-12 | 2000-09-29 | Nippon Telegr & Teleph Corp <Ntt> | 意味辞書登録装置 |
JP2010231433A (ja) * | 2009-03-26 | 2010-10-14 | Fujitsu Ten Ltd | 検索装置 |
Non-Patent Citations (2)
Title |
---|
See also references of EP2639705A4 * |
YU HIRATE ET AL.: "Keyword correction algorithm in keyword based search engine", DAI 2 KAI FORUM ON DATA ENGINEERING AND INFORMATION MANAGEMENT-DEIM 2010-RONBUNSHU, 9 June 2010 (2010-06-09), pages 1 - 8 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2014106832A (ja) * | 2012-11-29 | 2014-06-09 | Nec Software Tohoku Ltd | 情報提供装置及び情報提供方法 |
CN106407764A (zh) * | 2016-09-30 | 2017-02-15 | 深圳天珑无线科技有限公司 | 信息生成方法及装置 |
Also Published As
Publication number | Publication date |
---|---|
JP2013008372A (ja) | 2013-01-10 |
EP2639705A4 (en) | 2014-01-08 |
CN103279557A (zh) | 2013-09-04 |
KR20130080059A (ko) | 2013-07-11 |
CN103279557B (zh) | 2016-08-17 |
EP2650805A2 (en) | 2013-10-16 |
EP2650805B1 (en) | 2017-08-30 |
US20130346391A1 (en) | 2013-12-26 |
CA2822273A1 (en) | 2012-05-18 |
BR122013013420A2 (pt) | 2019-08-06 |
KR20130083468A (ko) | 2013-07-22 |
BR112013011573B1 (pt) | 2021-01-12 |
BR122013013420B1 (pt) | 2020-11-10 |
US8738366B2 (en) | 2014-05-27 |
ES2577938T3 (es) | 2016-07-19 |
KR101368594B1 (ko) | 2014-02-27 |
JP5101759B2 (ja) | 2012-12-19 |
BR112013011573A2 (pt) | 2016-08-09 |
ES2642379T3 (es) | 2017-11-16 |
US20130226563A1 (en) | 2013-08-29 |
EP2650805A3 (en) | 2014-01-08 |
KR101361403B1 (ko) | 2014-02-11 |
CA2822273C (en) | 2014-09-23 |
US8606565B2 (en) | 2013-12-10 |
EP2639705B1 (en) | 2016-04-13 |
CN103201737B (zh) | 2016-06-29 |
CA2817131A1 (en) | 2012-05-18 |
JP5170804B2 (ja) | 2013-03-27 |
CN103201737A (zh) | 2013-07-10 |
CA2817131C (en) | 2014-08-26 |
JPWO2012063770A1 (ja) | 2014-05-12 |
EP2639705A1 (en) | 2013-09-18 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP5170804B2 (ja) | 関連語登録装置、情報処理装置、関連語登録方法、関連語登録装置用プログラム、および、記録媒体 | |
US10783156B1 (en) | Scoring candidate answer passages | |
US8001135B2 (en) | Search support apparatus, computer program product, and search support system | |
US10691679B2 (en) | Providing query completions based on data tuples | |
US9990442B2 (en) | Method for determining relevant search results | |
US9317606B1 (en) | Spell correcting long queries | |
JP6506489B1 (ja) | 特許評価判定方法、特許評価判定装置、および特許評価判定プログラム | |
EP2720156B1 (en) | Information processing device, information processing method, program for information processing device, and recording medium | |
JP2017045196A (ja) | 曖昧性評価装置、曖昧性評価方法、及び曖昧性評価プログラム | |
TWI529546B (zh) | Information processing apparatus, information processing method and recording medium | |
JP4839295B2 (ja) | クエリ抽出方法、クエリ抽出装置およびクエリ抽出プログラム | |
JP2007188134A (ja) | 索引ファイルを用いた文書検索の方法 | |
WO2012052983A1 (en) | Method for scoring and ranking search engine keywords at a website | |
JP2020021455A (ja) | 特許評価判定方法、特許評価判定装置、および特許評価判定プログラム | |
JP2013109514A (ja) | 関連ワード表示制御装置、関連ワード表示方法、及びプログラム | |
JP2009277154A (ja) | 情報検索装置及び情報検索方法 |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 11839828 Country of ref document: EP Kind code of ref document: A1 |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2012533417 Country of ref document: JP |
|
ENP | Entry into the national phase |
Ref document number: 2817131 Country of ref document: CA |
|
WWE | Wipo information: entry into national phase |
Ref document number: 13884344 Country of ref document: US |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
ENP | Entry into the national phase |
Ref document number: 20137014718 Country of ref document: KR Kind code of ref document: A |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2011839828 Country of ref document: EP |
|
REG | Reference to national code |
Ref country code: BR Ref legal event code: B01A Ref document number: 112013011573 Country of ref document: BR |
|
ENP | Entry into the national phase |
Ref document number: 112013011573 Country of ref document: BR Kind code of ref document: A2 Effective date: 20130509 |