JP2009230173A - Synonym conversion system, synonym conversion method and synonym-converting program - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To convert a term having different concepts by situations such that the term has a concept in line with a situation. <P>SOLUTION: A synonym conversion system includes: a concept dictionary storage means for storing a concept dictionary wherein the concepts of the situation change type term having the different concepts by the situations are defined by the situations; a target term sorting means for sorting a term included in designated information into the situation change type term that is a conversion target and a term except the term; and a concept acquisition means for retrieving the situation change type term sorted by the target term sorting means from the concept dictionary by using the non-sorted other terms as information to a situation item of the term that is the conversion target in the concept dictionary to obtain the concept associated with the term that is the conversion target. <P>COPYRIGHT: (C)2010,JPO&INPIT

Description

  The present invention relates to a synonym conversion system, a synonym conversion method, and a synonym conversion program for converting a word into another word according to a concept of a word.

  Regarding synonym conversion technology, for example, Patent Document 1 describes a text-to-speech method in which a part expressing a quantity is detected, and when an expression that paraphrases the part is set, the expression is changed to another expression.

  Further, for example, Patent Document 2 describes an information processing apparatus that reads and converts a corresponding standard word from a dialect dictionary.

JP 09-198389 A JP 2005-331608 A

  In synonym conversion in which a word is converted into another word in accordance with the concept of a word, it is important to grasp the synonym relationship between words. For example, it can be said that the age of “13 years old” is generally “child”, but may be treated as “junior high school student” or “adult” depending on the situation. That is, the numeral “13 years old” has a plurality of concepts of “children”, “junior high school students”, and “adults”, and it can be said that they are synonymous with these, but such synonym relations are difficult to recognize, There was a problem that proper conversion was not performed.

  For example, if the method described in Patent Document 1 is used, the quantity can be converted into a different expression. However, the method described in Patent Document 1 is a conversion method based on a unique synonym relationship that is simply converted into a set expression, and the concept that one quantity has depends on the situation. No consideration is given. Therefore, it is not possible to convert the concept according to the situation as appropriate.

  For example, if the information processing apparatus described in Patent Document 2 is used, dialects that differ for each region can be converted into one standard language. However, the method described in Patent Document 2 is to convert one common concept word that is expressed differently in each region into a standard word, and there are also words that have different concepts depending on the situation. Is not considered at all. Therefore, it is not possible to convert the concept according to the situation as appropriate by simply using a dialect dictionary that registers synonyms by region. For example, it is impossible to obtain different conversion results for each user depending on the user's situation.

  Accordingly, an object of the present invention is to provide a synonym conversion system, a synonym conversion method, and a synonym conversion program capable of converting words having different concepts depending on the situation with a concept that matches the situation. To do.

  A synonym conversion system according to the present invention is a synonym conversion system that converts a word into another word in accordance with the concept of a word, and the word is converted to a context-changing word having a different concept depending on the situation. A concept dictionary storage means for storing a concept dictionary that associates information indicating the concept of each situation according to the situation, and selects words included in the specified information as context change type words to be converted and words that are not The target word selecting unit and the concept using the other words that have not been selected as information on the status item of the word to be converted in the concept dictionary for the status change type words selected by the target word selecting unit It comprises a concept acquisition means for acquiring a concept associated with the word to be converted by searching a dictionary.

  Further, the synonym conversion method according to the present invention is a synonym conversion method for converting a word into another word according to the concept of a word, and converts the word into a status change type word having a different concept depending on the situation. On the other hand, a concept dictionary in which information indicating the concept of the word is associated with each situation is stored, and words included in the specified information are classified into a context-changing word to be converted and a word that is not. Then, with respect to the selected status change type word, by searching the concept dictionary using other words that have not been selected as information on the status item of the word to be converted in the concept dictionary, the conversion target It is characterized by acquiring a concept associated with the word.

  A synonym conversion program according to the present invention is a synonym conversion program for converting a word into another word according to a concept of a word, and is a situation change type having a concept that varies depending on the situation. A computer having storage means for storing a concept dictionary in which information indicating the concept of the word is associated with each word for each word. Target word selection process for selecting words that are not to be used, and other words that have not been selected for the selected status change-type words, as information on the status items of the words to be converted in the concept dictionary A concept acquisition process for acquiring a concept associated with the word to be converted is executed by searching the concept dictionary.

  According to the present invention, words having different concepts depending on the situation can be converted into concepts that match the situation.

  Embodiments of the present invention will be described below with reference to the drawings. FIG. 1 is a block diagram showing a configuration example of a synonym conversion system according to the present invention. The synonym conversion system shown in FIG. 1 includes a concept dictionary storage unit 1, a target word selection unit 2, and a concept acquisition unit 3.

  The concept dictionary storage unit 1 stores a concept dictionary in which information indicating the concept of the word is associated with each situation for each situation change type word having a concept that varies depending on the situation. For example, in the case of using a number as a type of a context change type word to be converted, the concept dictionary storage unit 1 limits the use of the number and uses a numerical value that the number can take. You may memorize | store the concept dictionary which matched the information which shows the concept which a numerical numeral number has according to the situation.

  The target word selection unit 2 selects words included in the designated information into a status change type word to be converted and a word that is not. Note that “words included in specified information” are not limited to words included as character strings, but are words indicated by data (numerical data registered by item, etc.) included in specified information. Including. For example, the target word selection unit 2 may perform the selection by determining whether or not the word included in the specified information is a numeral. In addition, as shown in FIG. 2, when the system includes a target word definition table storage unit 4 that defines a status change type word to be converted, the target word selection unit 2 is designated. The selection may be performed by determining whether or not the word included in the information is a word defined in the target word definition table stored in the target word definition table storage unit 4. FIG. 2 is a block diagram showing another configuration example of the synonym conversion system according to the present invention.

  The concept acquisition unit 3 uses the concept dictionary by using other words that have not been selected as information on the status items of the words to be converted in the concept dictionary for the status change type words selected by the target word selection unit 2. By searching, a concept associated with the conversion target word is acquired.

  Further, FIG. 2 shows an example in which the situation data input means 5 is further provided. The situation data input means 5 inputs situation data indicating the situation of a certain thing. In such a case, the target word selection unit 2 selects the words included in the situation data input by the situation data input unit 5 into the status change type words to be converted and the words that are not. Good. Further, the target word selection unit 2 is, for example, information on the situation data input by the situation data input unit 5 or an object to be the target of the situation data, and is information stored in advance and without short-term fluctuation. The words included in the profile data may be classified into a status change type word to be converted and a word that is not.

  Further, the situation data input means 5 may input situation data indicating the current situation of the target thing according to a predetermined timing. Moreover, you may input the situation data which shows the condition selected by the user as situation data which shows the condition of a target thing.

  Further, FIG. 2 shows an example in which the situation data extracting means 6 is further provided. The situation data extraction means 6 extracts the situation data by processing the words that have not been selected as the status change type words to be converted by the target word selection means 2 by a predetermined method. The situation data extraction means 6 is means for automatically obtaining situation data from words used together with the situation change type words to be converted, and corresponds to the situation items used in the concept dictionary. A means for obtaining status data in a format. For example, the situation (for example, the position of the target object) is obtained by performing morphological analysis on the words that are not selected as the status change type words to be converted by the target word selection unit 2 and decomposing the compound words. The situation data indicating the location may be extracted. In addition, for example, by converting using a conversion table, situation data expressing a situation indicated by a word that is not a situation change type to be converted in a format corresponding to a situation item used in the concept dictionary may be extracted. . Note that the situation data extraction unit 6 is not necessary when the situation data input unit 5 inputs situation data according to the purpose in a format corresponding to the situation item used in the concept dictionary.

  Hereinafter, more specific embodiments will be described.

Embodiment 1. FIG.
FIG. 3 is a block diagram showing a configuration example of the synonym conversion system as the first embodiment of the present invention. The synonym conversion system shown in FIG. 3 includes a conversion target data input unit 101, a situation data input unit 102, a selection unit 201, a situation data extraction unit 202, a situation correspondence type concept acquisition unit 203, and an output unit 4. Is provided. In addition, storage means for storing the target word definition table 301 and the concept dictionary 302 is provided.

  The conversion target data input unit 101 inputs data to be subjected to synonym conversion of the present system (hereinafter referred to as conversion target data). The conversion target data may be any information. For example, it may be information about a certain person. The conversion target data input unit 101 inputs conversion target data in accordance with, for example, a key operation by an operator. It is also possible to omit the conversion target data input unit 101 and store the conversion target data in the storage unit in advance.

  The situation data input means 102 inputs data indicating the situation of an object indicated by the conversion target data (hereinafter referred to as situation data). The situation data input means 102 inputs situation data according to, for example, a key operation by an operator. In addition, for example, by realizing with a timing device, a GPS receiver, a measuring instrument, an RFID reader, etc., time information indicating the current situation of the thing indicated by the conversion target data, position information, physical information, climate information, etc. You may enter. When the situation data input unit 102 inputs a plurality of data, one of them can be used as conversion target data.

  The sorting unit 201 sorts the words included in the designated information into the status change type words (words having different concepts depending on the situation) to be converted and the words that are not. The selection unit 201 may select, for example, by determining whether or not a part of speech (for example, a number) is defined in advance. In the present embodiment, the selection is performed by referring to a target word definition table 301 to be described later and determining whether the word is defined in the target word definition table 301. Note that the words selected by the selection unit 201 become the processing target words of the situation corresponding concept acquisition unit 203.

  The target word definition table 301 is a table that defines a status change type word to be converted. For example, in the case of defining a numerical word as a situation-changing word, a list of words that come immediately before a number such as a classifier or unit may be registered in combination with a special symbol indicating an arbitrary number.

  The situation data extraction unit 202 processes the words included in the status data and the words selected in the status change type words that are not converted by the selection unit 201 among the words included in the conversion target data, and further the situation Extract data. For example, the situation data extraction unit 202 may extract situation data that matches a data format used as a situation item in the concept dictionary 302 described later by performing morphological analysis, latitude / longitude conversion, or the like. Note that the situation data input means 102 and the situation data extraction means 202 are complementary to each other in that situation data is acquired, and either of them can be omitted.

  The situation-corresponding concept acquisition unit 203 acquires a concept that matches the situation indicated by the situation data for the situation-changing type word selected as the conversion target by the selection unit 201. The situation-corresponding concept acquisition unit 203 searches a concept dictionary 302 described later, and associates the situation-changing word selected as a conversion target in the concept dictionary 302 with a situation that matches the situation indicated by the situation data. Get the concept.

  The output unit 4 outputs the concept acquired by the situation corresponding type concept acquisition unit 203.

  Next, the operation of this embodiment will be described. FIG. 4 is a flowchart showing an example of the operation of this embodiment. As shown in FIG. 4, first, the conversion target data input unit 101 inputs conversion target data to be subjected to synonym conversion of this system (step S101). Next, the status data input means 102 inputs status data indicating the status of the thing indicated by the conversion target data (step S102). If the status data input means 102 is not operated in conjunction with the conversion target data input means 101, such as when the conversion target data is stored in advance, it is driven periodically or triggered by a user action. The situation may be acquired.

  Next, the screening unit 201 refers to the target word definition table 301 and selects a status change type word to be converted from the specified information (in this case, conversion target data) (step S103).

  Next, the situation data extraction unit 202 further extracts situation data from words other than the word to be converted (step S104). It should be noted that the words included in the situation data are also included in the words other than the word to be converted.

  Then, the situation corresponding concept acquisition unit 203 searches the concept dictionary 302 for the situation change type word selected as the conversion target, and uses the situation data acquired in step S102 and step S104 among the concepts of the word. A concept associated with the indicated situation is acquired (step S105).

  Finally, the output unit 4 outputs the concept acquired by the situation-corresponding concept acquisition unit 203 in association with the situation change-type word as the target word (step S106). The situation data used for the determination may also be output.

  Hereinafter, the operation of this embodiment will be described with a specific example. In this example, a GPS (Global Positioning System) receiver is used as the status data input means 102. Further, a magnetic disk is used as the storage means. A display device is used as the output means. A CPU that operates according to a program is used as a data processing device that implements the functions of the selection unit 201, the situation data extraction unit 202, and the situation corresponding concept acquisition unit 203. In addition, the magnetic disk of this example includes not only a number definition table (target word definition table 301) in which a number to be converted is defined, but also a number of conversion target data in addition to a number dictionary contextual concept dictionary (concept dictionary 302). A user profile as a section, a situation storage database for accumulating situation data, and a machining table for machining are stored.

  The status data input means 102 is driven periodically or triggered by a user action, and acquires the current status of the user. As user status, for example, (1) time information such as date, day of the week, time, etc., (2) location information such as address, longitude / latitude, store name, station name, etc., (3) body such as weight, blood pressure, pulse, etc. Information (4) Acquire and input climate information such as temperature, humidity, and wind power. It should be noted that a plurality of situation data can be input by mounting a plurality of situation data input means 102.

  For example, when the situation data input unit 102 is realized by a GPS receiver, the user's position information may be periodically input. Further, for example, when the situation data input unit 102 is realized by a network communication unit, user position information (position coordinate information, station name, etc.) captured by a center device in an RFID system or the like may be input. Further, for example, when the user passes the credit card through the card reader, the user's position information (store information) captured by the center device may be input.

  For example, when the situation data input unit 102 is realized by a measuring device such as a weight scale or a sphygmomanometer, the body information may be input together with the time information when the user operates the measuring device. For example, when the situation data input unit 102 is realized by a measuring instrument that measures climate information such as temperature, humidity, and wind power, the information may be periodically input. For example, when the status data input unit 102 is realized by an input device such as a keyboard or a mouse, information specified by the user may be input in response to a user operation.

  Further, the situation data input by the situation data input means 102 is registered in the situation storage database. When a plurality of types of situation data are input, they are stored for each item so as to know what information. For example, the blood pressure status acquired by the sphygmomanometer is registered in the item “blood pressure”. Moreover, what is necessary is just to memorize | store for every user, when providing a synonym conversion service with respect to several users.

  FIG. 5 is an explanatory diagram showing an example of holding status data. In the example shown in FIG. 5, status data is held for each user and for each item. In FIG. 5, for example, time information (year / month / day, day of the week, time) = 13: 15 on Tuesday, January 1, 2008 and position = Shibaura movie theater are stored as status data associated with the user A. An example is shown. Further, for example, as the situation data associated with the user B, an example is shown in which weight = 50 kg, blood pressure = 110 mmHg, and pulse = 82 bpm are stored. In FIG. 5, the items illustrated for the user A and the user B are different from each other.

  In the user profile, matters with no short-term fluctuation are registered for the user. FIG. 6 is an explanatory diagram illustrating an example of a user profile for the user A. In the example shown in FIG. 6, a user profile indicating that the user A is 13 years old, the gender is male, and the height is 150 cm is shown.

  A list of classifiers and units used in combination with the numbers to be converted in this system is registered in the number definition table. FIG. 7 is an explanatory diagram showing an example of the number definition table. In addition, as shown in FIG. 7, you may make it register the thing including the variation about what uses several expression per unit. Moreover, you may assign an identifier (ID) for every kind of number.

  In this example, the selection unit 201 operates in conjunction with the situation data input unit 102. When the situation data input unit 102 inputs the user's situation, the selection unit 201 reads the information about the user from the situation storage database and the user profile, refers to the number definition table, and each data included in the information about the user is The words to be shown are sorted into the number to be converted and the other words. That is, in this example, the user profile and the user status data that is the target of the user profile stored in the status storage database are used as conversion target data and status data without any particular distinction.

  The selecting unit 201 inputs information on the words selected as the number to be converted into the situation-corresponding concept acquisition unit 203 and inputs information on other words into the situation data extracting unit 202. Note that the selection unit 201 can also select and determine whether or not a word that is a unit is followed by a number without referring to the numeral definition table. It is also possible to select the words by omitting the number definition table and determining whether or not the words are registered in the concept dictionary 302.

  Also, rules and correspondences for processing words other than numerals into status data are registered in the processing table. For example, a rule for performing morphological analysis and a correspondence relationship (correspondence between store name and latitude and longitude) for obtaining latitude and longitude from the store name are registered.

  The situation data extraction unit 202 processes the word input from the selection unit 201 into situation data with reference to a processing table as necessary, and inputs the situation data to the situation correspondence type concept acquisition unit 203. Here, the processing specifically refers to obtaining situation data by performing morphological analysis and decomposing a compound word, or converting a store name into latitude and longitude as situation data.

  Further, in the concept dictionary 302, for each number whose concept changes depending on the situation, a concept for each situation that the value has is registered. In this example, the correspondence between the number, the situation, and the concept is formed as a set of dictionaries collected for each category (type of number). FIG. 8 is an explanatory diagram illustrating an example of the concept dictionary 302. FIG. 8 shows an example of a dictionary in which the concept according to the situation about age is registered. As shown in FIG. 8, for example, in the category of age, the age from the age of 0 is arranged in the heading, and the concept in each status item such as “movie theater”, “train”, and “bus” is associated with the horizontal axis. In the example illustrated in FIG. 8, the concept of conversion when the situation is not particularly defined is associated with the item “general”. For example, “junior high school student” is associated with the age “13 years old” as a concept when the situation is “movie theater”. This indicates that the age of 13 is treated as a junior high school student in a movie theater. In the concept dictionary 302, not only “notice” such as “children” and “children”, “junior high school” and “junior high school students”, but also “children” and “children”, “elderly” and “senior” have the same meaning. It is preferable that words with different expressions can be handled centrally. It is also possible to indicate a numerical value to be registered as a heading by specifying a range.

  The situation-corresponding concept acquisition means 203 searches the concept dictionary 302 using a keyword other than a number as the situation data input from the selection means 201 and the situation data input from the situation data extraction means 202 as a keyword. The concept in the situation indicated by the situation data is acquired and output to the output means 4. It is assumed that the situation-responsive concept acquisition unit 203 also obtains information about what the numeral input from the selection unit 201 indicates. The situation corresponding concept acquisition unit 203 first checks whether or not the number input from the selection unit 201 exists in the heading of the concept dictionary 302. When there is a matching headline, it is checked whether or not a word other than a number as the status data input from the status data extraction unit 202 exists in the status item associated with the heading. If there is a matching situation item, the heading and the concept associated with the situation item are acquired as the concept of the number.

  The output unit 4 displays the concept acquired by the situation corresponding concept acquisition unit 203. The content to be displayed may be only the character string of the concept, or may be displayed together with the number and the situation that are the basis for obtaining the concept. The output unit 4 may output a control signal including information on the acquired concept to a unit that uses the acquired concept.

  For example, the movement of the user A is periodically captured by GPS, and the situation at a certain point in time is “Year / Month = January 1, 2008”, “Day of the week = Tuesday”, “Time = 13: 15”, “Position = Shibaura Movie” Suppose you get situation data. It is assumed that data “age = 13 years old”, “sex = male”, and “height = 150 cm” are registered in advance in the user profile. Further, it is assumed that a word in which at least “year” follows a number is defined as a target word in the numeral definition table.

  Here, when the GPS is activated and the status data is input, the sorting unit 201 is activated. The selection unit 201 uses the status storage database and the user profile as information on the user A as “date of year = January 1, 2008”, “day of the week = Tuesday”, “time = 13: 15”, “position = Shibaura Movie Theater”. ”,“ Age = 13 years old ”,“ sex = male ”, and“ height = 150 cm ”, and these are matched with the number definition table (comparison determination is performed by comparison). Here, “13 years old” is selected as a target word. Then, the information of the word is output to the situation corresponding concept acquisition means 203. In addition, the sorting means 201 includes other words “2008”, “January”, “1 day”, “Tuesday”, “13:00”, “15 minutes”, “Shibaura movie theater”, “male” ”And“ 150 cm ”are output to the situation data extraction unit 202.

  The situation data extraction unit 202 performs morphological analysis on “Tuesday”, “Shibaura Movie Theater”, and “Men” in the input information to obtain “Tuesday”, “Sun”, “Shibaura”, “Movie Theater”, and “Men”. Then, they are output to the situation corresponding concept acquisition means 203 as situation data. The situation data extraction unit 202 is given a category to be extracted as situation data in advance. For example, only “movie theaters” that match facility information may be output as situation data in this system.

  The situation corresponding concept acquisition unit 203 first checks whether or not the target word input from the selection unit 201 exists in the heading of the concept dictionary 302. Here, “13 years old” matches in the category of age. Next, whether or not the words “Tuesday”, “Sunday”, “Shibaura”, “Movie Theater”, and “Men” other than the numbers entered as situation data exist in the status items associated with the age category Check. Here, it is assumed that “movie theater” matches. Then, the situation corresponding concept acquisition means 203 acquires the concept “junior high school student” associated with the headline “13 years old” in the situation item “movie theater” as the concept of the number “13 years old”. The output means 4 outputs “junior high school student” as a concept corresponding to “13 years old” in the “movie theater”. Thereby, when the user A pays an entrance fee in a movie theater, it turns out that the fee as a junior high student should just be paid.

  As described above, according to the present embodiment, a concept that matches the situation associated with the number (here, the situation of the user A representing the age) is associated with the number having a different concept depending on the situation. Can be sought. For example, the age of “13 years old” is treated as “junior high school student” in “cinema” and as “adult” in “bus” so that it can be easily understood without checking in the age table each time. Become.

Embodiment 2. FIG.
Next, a second embodiment of the present invention will be described. FIG. 9 is a block diagram showing a configuration example of a synonym conversion system as the second exemplary embodiment of the present invention. The synonym conversion system shown in FIG. 9 is different from the first embodiment shown in FIG. 3 in that it further includes a sentence creation unit 221. In addition, storage means for storing the sentence creation rule 321 is provided.

  The sentence creation rule 321 registers a rule for converting the concept acquired by the situation-based concept acquisition unit 203 into a sentence. In this embodiment, a sentence example expressing the concept of the target word is registered for each category of the target word that is the basis for obtaining the concept. Furthermore, sentence examples may be registered for each status item. Note that the sentence example includes a special symbol for specifying a location where the concept is used in the sentence example.

  The sentence creation unit 221 receives the concept acquired by the situation-corresponding concept acquisition unit 203, and creates a sentence using a word expressing the concept based on the sentence creation rule 321.

  In the present embodiment, the output unit 4 outputs the text created by the text creation unit 221.

  Next, the operation of this embodiment will be described. FIG. 10 is a flowchart showing an example of the operation of the present embodiment. Note that steps S101 to S105 shown in FIG. 10 are the same as those in the first embodiment shown in FIG.

  In the present embodiment, when the situation corresponding concept acquisition unit 203 acquires the concept in the situation indicated by the situation data for the word selected as the target word (Step S105), the acquired concept is output to the sentence creation unit 221.

  The sentence creation unit 221 creates a sentence using a word expressing the concept acquired by the situation corresponding concept acquisition unit 203 based on the sentence creation rule 321 (step S201). For example, the sentence creation unit 221 may extract a sentence example registered in the sentence creation rule 321 and apply a concept acquired to the sentence example to create a sentence expressing the concept.

  Finally, the output unit 4 outputs the sentence created by the situation corresponding concept acquisition unit 203 (step S202).

  Hereinafter, the operation of this embodiment will be described with a specific example. This embodiment is an example in which temperature data is used as data to be converted, and sentences are output using a concept corresponding to an observation position associated with the temperature data. In this example, a thermometer is used as the status data input means 102 in addition to the GPS receiver. In this example, temperature data input by a thermometer which is the situation data input unit 102 is handled as conversion target data, and an example is shown in which the situation data input unit 102 also functions as the conversion target data input unit 101. Yes.

  Note that the sentence creation unit 221 is realized by a CPU that operates according to a program, like other processing units. Moreover, the point from which the text creation rule 321 is memorize | stored in the magnetic disc as a memory | storage means differs from 1st Embodiment. In this embodiment, it is assumed that the thermometer measures the Fahrenheit temperature.

  Now, assume that the thermometer is operated and information indicating the temperature of Fahrenheit “33 degrees” is input. Here, it is assumed that the GPS receiver inputs information indicating the current position “America” in conjunction with the thermometer.

  In this embodiment, the selection unit 201 determines the Fahrenheit temperature “33 ° C.” as a target word and inputs it to the situation corresponding concept acquisition unit 203.

  If the GPS receiver cannot obtain the situation data in the format indicating the country name, the situation data extraction unit 202 may extract the country name “USA” used for the situation item from the latitude and longitude data.

  Then, the situation corresponding concept acquisition unit 203 obtains “cold” as a concept corresponding to the situation data “America” for the number “33 degrees” representing the temperature determined as the target word. The operation so far can be performed with the same configuration as that of the first embodiment.

  Next, the sentence creation means 221 extracts from the sentence creation rule 321 the sentence example “It is cold” associated with the category “temperature” of the target word that is the basis for obtaining the concept, and the sentence example. Apply a word that expresses the concept acquired in step 1, and create a sentence that says "It's cold". In the above example, the portion surrounded by “” is registered as a portion using a word expressing the acquired concept.

  The output means 4 outputs the prepared sentence “It is cold”. The output means 4 may use a display device or a speaker. The output sentence may be used for creating a letter sentence or may be used for automatic conversation by a robot or the like.

  When a plurality of sentence examples are registered, a sentence corresponding to each of the sentence examples may be created and displayed, and the user may select it.

  As described above, according to the present embodiment, words having a different concept depending on the situation such as a number can be documented using the concept that matches the situation at that time.

Embodiment 3. FIG.
Next, a third embodiment of the present invention will be described. FIG. 11 is a block diagram showing a configuration example of a synonym conversion system as the third exemplary embodiment of the present invention. The synonym conversion system shown in FIG. 11 is different from the first embodiment shown in FIG. 3 in that the conversion target data input means 101 is the search request input means 131 and the situation data input means 102 is the input status. The difference is that the data input unit 132 is provided, and the search unit 231 and the filtering unit 232 are provided.

  In the present embodiment, information related to a user who has input a search request is set as conversion target data and situation data, and the concept of the target word included in the conversion target data (here, the result of conversion to a synonym and agreement) is searched. Used for filtering on results.

  The search request input means 131 inputs a search request from the user. The input status data input unit 132 operates in conjunction with the search request input unit 131, and inputs information related to the user who has input the search request as conversion target data and status data indicating the status related to the search request.

  In addition, the search unit 231 searches a predetermined content database based on a search request input via the search request input unit 131, and obtains content information that matches the conditions as a search result. The filtering unit 232 narrows down the search result by the search unit 231 using the concept acquired by the situation-based concept acquisition unit 203 from the information about the user who has input the search request as a narrowing condition.

  The output unit 4 outputs the search result narrowed down by the filtering unit 232.

  Next, the operation of this embodiment will be described. FIG. 12 is a flowchart showing an example of the operation of the present embodiment. In this embodiment, first, the search request input unit 131 accepts a search request from the user (step S301).

  When the search request input unit 131 receives the search request, the search unit 231 executes a search process on the content database based on the received search request, and obtains information on some contents as a search result (step S321). ).

  On the other hand, when the search request input unit 131 receives the search request, the input status data input unit 132 inputs information about the user who has input the search request (step S311). The process of step S321 is also realized by, for example, reading out a separately stored user profile or situation storage database by allowing the user to input information for identifying the user as a pre-process for inputting a search request. May be.

  Next, the selecting unit 201 refers to the target word definition table 301 and uses the specified information (in this case, information about the user input by the input status data input unit 132) to convert the status change type to be converted. A word is selected (step S312).

  Next, the situation data extraction unit 202 extracts situation data from words other than the word to be converted (step S313). Note that if the word already included in the information related to the user is in a format that matches the status item used in the concept dictionary 302, the process of step S313 may be omitted. In such a case, words other than the word to be converted may be directly input to the situation corresponding concept acquisition unit 203 as situation data.

  Next, the situation correspondence type concept acquisition unit 203 searches the concept dictionary 302 for the situation change type words selected as the conversion target, and the situations acquired in step S311 and step S313 among the concepts of the words. A concept associated with the situation indicated by the data is acquired (step S314).

  In step S315, the filtering unit 232 narrows down the search result obtained in step S321 using the concept acquired in step S314. For narrowing down, not only the concept of the situation-changing word to be converted but also the concept of another word may be used. It is also possible to prioritize the items to be used and narrow down while increasing the number of items until a predetermined number of items are reached.

  Finally, the output unit 4 displays the search result narrowed down by the filtering unit 232 (step S322).

  Hereinafter, the operation of this embodiment will be described with a specific example. In this example, the search request input unit 131 and the input status data input unit 132 are realized by an input device such as a mouse or a keyboard. Note that the input status data input means 132 can be realized by a measuring device, a card reader, or the like, similarly to the status data input means 102 in the first embodiment.

  The search request received by the search request input unit 131 may be a word, a sentence, or an expression. Moreover, one or more may be sufficient. In this example, a case where a search keyword is used as a search request is described as an example.

  Now, it is assumed that the user C inputs a search keyword “potato onion apple menu” via the search request input means 131 for the purpose of searching for a menu from the Web site.

  The search unit 231 searches the predetermined content database for content including all these four words, and obtains information on some content (Web pages) as a search result. Alternatively, information that there is no corresponding content may be obtained.

  Further, when the search request input unit 131 inputs a search keyword, the input situation data input unit 132 operates, and information on the user C when the search request is input is acquired as conversion target data and situation data. Note that this example is realized by acquiring information for identifying the user C who is an input person and reading the latest information from a database in which information on the user C is regularly updated. Here, it is assumed that data of “sex = female” and “blood pressure = 128 mmHg” are acquired.

  Then, the selection unit 201 determines the blood pressure “128 mmHg” as a target word and inputs it to the situation corresponding concept acquisition unit 203. The situation corresponding concept acquisition unit 203 obtains “high blood pressure” as a concept corresponding to “female” indicated by the situation data for the number “128 mmHg” representing the blood pressure determined as the target word. The operation so far can be performed with the same configuration as that of the first embodiment.

  In the present embodiment, the filtering unit 232 then narrows down the search results by the search unit 231 using the concept of “high blood pressure” acquired by the situation corresponding concept acquisition unit 203.

  Then, the output unit 4 outputs the search result narrowed down by the filtering unit 232. The output unit 4 may display, for example, a title list of content finally obtained as a search result. The content to be displayed is not limited to the title but may be a URL of the content or a summary. Moreover, you may combine them.

  As described above, according to the present embodiment, when the search result is narrowed down based on the information about the user who has input the search request, the information includes words having different concepts depending on the situation such as a number. Even if it is, it can narrow down appropriately. Therefore, it is possible to provide information more in line with the current user situation from information desired by the user.

  In the above description, an example is shown in which the search result is filtered. However, by adding a condition to the search request, the search result narrowed down from the beginning may be acquired. Moreover, although the example which uses only the information regarding a user as conversion object data is shown, a search request | requirement may be included in conversion object data in consideration of the case where a number etc. are contained in a search request. In such a case, content that does not hit only by the number alone by adding a concept that matches the user information of the word as an OR condition to the state change type word such as a number included in the search request. It is also possible to search for.

  The present invention is not limited to a system whose purpose is to convert a word into a synonym, but a synonym relationship can be grasped by a different concept depending on the situation, such as a system for generating a sentence and a system for filtering information. In a system in which a difficult word is input, it is possible to make the word easy to use by making the meaning of the word unique. Therefore, any system that uses a phrase for processing can be preferably used.

It is a block diagram which shows the structural example of the synonym conversion system by this invention. It is a block diagram which shows the other structural example of the synonym conversion system by this invention. It is a block diagram which shows the structural example of the synonym conversion system as the 1st Embodiment of this invention. It is a flowchart which shows an example of operation | movement of 1st Embodiment. It is explanatory drawing which shows the holding example of status data. It is explanatory drawing which shows an example of a user profile. It is explanatory drawing which shows an example of a number definition table. It is explanatory drawing which shows an example of the concept dictionary. It is a block diagram which shows the structural example of the synonym conversion system as the 2nd Embodiment of this invention. It is a flowchart which shows an example of operation | movement of 2nd Embodiment. It is a block diagram which shows the structural example of the synonym conversion system as the 3rd Embodiment of this invention. It is a flowchart which shows an example of operation | movement of 3rd Embodiment.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Concept dictionary memory | storage means 2 Object word selection means 3 Concept acquisition means 4 Object word definition table storage means 5 Situation data input means 6 Situation data extraction means 7 Utilization means 101 Conversion object data input means 102 Situation data input means 201 Selection means 202 Situation Data extraction means 203 Situation correspondence type concept acquisition means 301 Synonym definition table 302 Concept dictionary 4 Output means 221 Document creation means 321 Sentence creation rules 131 Search request input means 132 Situation data input means 231 Search means 232 Filtering means

Claims (14)

A synonym conversion system for converting a word into another word according to a concept of a word,
Concept dictionary storage means for storing a concept dictionary in which information indicating the concept of the word is associated with each situation for each situation change-type word having a different concept depending on the situation;
An object word selection means for selecting a word included in the specified information into a status change type word to be converted and a word not to be converted;
By searching the concept dictionary using other words that have not been selected as information on the status item of the word to be converted in the concept dictionary for the status change type words selected by the target word selection means And a synonym conversion system comprising: concept acquisition means for acquiring a concept associated with the word to be converted. Comprising a target word definition table storage means for storing a target word definition table that defines a context change type word to be converted;
The target word sorting unit performs sorting by determining whether or not the word included in the specified information is a word defined in the target word definition table stored in the target word definition table storage unit. The synonym conversion system according to claim 1. The concept dictionary storage means, when using a number as a type of a context-change-type word to be converted, limits the use of the number and uses the numerical value of the numerical value for the numerical value that the numerical value can take. The synonym conversion system of Claim 1 or Claim 2 which memorize | stores the concept dictionary which matched the information which shows the concept which has for every situation. A situation data input means for inputting situation data indicating the situation of a certain thing is provided.
The target word selection unit selects a word included in the situation data input by the situation data input unit into a status change type word to be converted and a word that is not to be converted. The synonym conversion system of any one of Claims. The target word selection means is information about the situation data input by the situation data input means or the information that is the object of the situation data, and is included in the profile data that is prestored information without short-term fluctuations. The synonym conversion system according to claim 4, wherein the words are classified into a status change type word to be converted and a word that is not. The synonym conversion system according to claim 4 or 5, wherein the status data input means inputs status data indicating the current status of the target object according to a predetermined timing. The synonym conversion system according to claim 4 or 5, wherein the situation data input means inputs situation data indicating the situation selected by the user as situation data indicating the situation of the target thing. A situation data extraction means for extracting situation data by processing a word that has not been selected as a status change type word to be converted by the target word selection means by a predetermined method;
The concept acquisition means searches the concept dictionary using the situation data extracted by the situation data extraction means as information on the situation item of the word to be converted in the concept dictionary, thereby obtaining the word to be converted. The synonym conversion system according to any one of claims 1 to 7, wherein a concept associated with is acquired. The synonym according to any one of claims 1 to 8, further comprising a utilization unit that uses a word representing one concept acquired by the concept acquisition unit as a synonym of the word to be converted. Conversion system. A synonym conversion method for converting a word into another word according to a concept of a word,
For a context-changing word having a different concept depending on the situation, a concept dictionary in which information indicating the concept of the word is associated with each situation is stored.
Select the words included in the specified information as status change words to be converted and words that are not,
For the selected status change type words, the other words that were not selected are used as information for the status items of the words to be converted in the concept dictionary, and are searched for the concept dictionary. A synonym conversion method characterized by acquiring a concept associated with a word. Stores the target word definition table that defines the status change type words to be converted,
The selection is performed by determining whether or not a word included in the specified information is a word defined in a target word definition table stored in the target word definition table storage unit. Synonym conversion method. In the case where a numer is defined as a type of a status change type word to be converted, the concept that the numerator of the numerical value has with respect to the numerical value that the numer can take after limiting the use of the numer The synonym conversion method of Claim 10 or Claim 11 which memorize | stores the concept dictionary which matched the information which shows according to a condition. A synonym conversion program for converting a word into another word according to a concept of a word,
A computer having storage means for storing a concept dictionary in which information indicating a concept of the word is associated with each situation for each situation change type word having a different concept depending on the situation,
Target word selection process for selecting words included in the specified information into context-change-type words to be converted and words that are not, and other words that were not selected for the selected status-change-type words A synonym for executing concept acquisition processing for acquiring a concept associated with the word to be converted by searching the concept dictionary using information as a status item of the word to be converted in the concept dictionary Word conversion program. In a computer provided with a storage means for storing a target word definition table that defines a context change type word to be converted,
In the target word selection process, selection is performed by determining whether or not the word included in the specified information is a word defined in the target word definition table stored in the target word definition table storage means. The program for synonym conversion according to claim 13.

Images