WO2008075524A1 - 極性推定システム、情報配信システム、極性推定方法及び、極性推定用プログラム、及び評価極性推定用プログラム - Google Patents
極性推定システム、情報配信システム、極性推定方法及び、極性推定用プログラム、及び評価極性推定用プログラム Download PDFInfo
- Publication number
- WO2008075524A1 WO2008075524A1 PCT/JP2007/072484 JP2007072484W WO2008075524A1 WO 2008075524 A1 WO2008075524 A1 WO 2008075524A1 JP 2007072484 W JP2007072484 W JP 2007072484W WO 2008075524 A1 WO2008075524 A1 WO 2008075524A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- polarity
- evaluation
- reputation information
- expression
- degree
- Prior art date
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q10/00—Administration; Management
- G06Q10/10—Office automation; Time management
Definitions
- Polarity estimation system information distribution system, polarity estimation method, polarity estimation program, and evaluation polarity estimation program
- the present invention relates to a polarity estimation system, a polarity estimation method, a polarity estimation program, and an evaluation polarity estimation program for estimating an evaluation polarity that indicates whether reputation information is positive or negative.
- the present invention relates to a polarity estimation system, a polarity estimation method, a polarity estimation program, and an evaluation polarity estimation program that estimate evaluation polarity by using reputation information for reputation information with unknown evaluation polarity.
- the present invention also relates to an information distribution system that distributes reputation information.
- the object is an object to be evaluated, and is, for example, a product name "X X PC" or a service name " ⁇ ⁇ service”.
- Reputation information is information that includes an expression of the content to be evaluated for the object, for example, information that includes an expression indicating the evaluation content such as “good”, “bad”, and “large”.
- the expression of the content for evaluating the object is called an evaluation expression.
- the reputation information may include an attribute expression indicating the attribute of the object! /.
- An attribute expression is a word indicating the characteristics of an object. For example, if a personal computer (hereinafter simply referred to as “PC”) is an object, it is a word such as “screen” or “weight”. is there.
- attribute expressions may be hierarchically connected. For example, from the input sentence (natural text text) that says “X X PC screen size is good”, the reputation information extraction system [object “O ⁇ PC”, attribute expression “screen”, Attribute expression “size”, evaluation expression “good” ] Reputation information is extracted.
- the object when a natural sentence text in which an object such as text on a bulletin board is obvious is input, the object is specified in the natural sentence text.
- the object may not be included in the reputation information that is not necessary.
- the reputation information when the attribute expression is omitted in the natural sentence text, the reputation information may not include the attribute expression.
- the reputation information may be a combination of the object, the attribute expression, and the evaluation expression.
- the reputation information may be a combination of the attribute expression and the evaluation expression, or a combination of the object and the evaluation expression. It ’s good!
- a reputation information extraction system is a system that inputs natural sentence text and extracts reputation information from the input natural sentence text.
- evaluation polarity is information indicating whether reputation information is positive or negative. For example, [object “O ⁇ PC”, attribute expression “screen”, attribute expression “size”, evaluation expression “good”] and! /, Reputation information is positive expression (in this example, expression “good” ), The evaluation polarity is positive. In the following, the evaluation polarity is simply the polarity! /.
- the evaluation polarity estimation system is a system that inputs reputation information and estimates the evaluation polarity of the input reputation information.
- each evaluation expression and the evaluation polarity for the evaluation expression are registered in a dictionary in advance, and the evaluation polarity of reputation information is estimated using this dictionary.
- the evaluation polarity estimation system described in Patent Document 1 includes an evaluation expression attribute storage unit, a negative expression storage unit, and evaluation expression attribute classification means.
- the evaluation expression attribute storage unit stores in advance a set of an evaluation expression and information indicating whether the evaluation expression is positive or negative.
- the negative expression storage unit stores negative expressions such as “no” and “no”.
- the evaluation expression attribute classification means classifies whether the reputation information is positive or negative.
- the evaluation expression attribute classification means inputs natural text text and position information indicating the appearance position of the evaluation expression. Then, the evaluation expression attribute classification means refers to the evaluation expression attribute storage unit, and classifies the reputation information as positive or negative according to the combination of the evaluation polarity of the evaluation expression and the negative expression around the evaluation expression. [0012] In addition, evaluation expressions often appear continuously in the text. Positive evaluation expressions are arranged before and after positive evaluation expressions, and negative evaluations are expressed before and after negative evaluation expressions. There tends to be many expressions. For example, Patent Document 2 describes a configuration configured to determine the evaluation polarity of reputation information using the hypothesis that there is such a tendency.
- the evaluation polarity estimation system described in Patent Document 2 includes a registered expression storage unit, an expression extraction unit, and a polarity determination unit.
- the registered expression storage unit stores in advance a set of an evaluation expression and information indicating whether the evaluation expression is positive or negative.
- the expression extraction unit extracts noun phrases and verb phrases from natural sentences.
- the polarity determination unit refers to the registered expression storage unit and determines that the verb phrase that appears along with the evaluation expression! /, Has the same evaluation polarity as the evaluation expression.
- the evaluation polarity estimation system described in Patent Document 2 when the evaluation polarity of! /, N, or verb phrase is registered in advance in the registered expression storage unit and exceeds a certain threshold, it is estimated to be the evaluation polarity.
- Patent Document 1 Japanese Patent Laid-Open No. 2002-92004 (Page 9, Fig. 9)
- Patent Document 2 Japanese Patent Laid-Open No. 2006-146567 (Page 9-10, Fig. 3)
- the evaluation polarity estimation system described in Patent Document 1 has a problem that it may be difficult to judge the evaluation polarity using only the evaluation expression.
- the evaluation expressions “like” and “excellent! /,” Can be judged as positive evaluation expressions, and “dislike! /,” And “first! /,” are negative evaluation expressions. I can judge. However, if the evaluation expression is “large”, it cannot be judged that it is generally a positive expression or a negative expression! /.
- reputation information object “computer”, attribute expression “screen”, evaluation expression “large”
- object “computer” Attribute expression“ noise ”, and evaluation expression“ large ”,“ large ”, negative reputation information and The Therefore, the evaluation polarity may not be determined only from the evaluation expression.
- the present invention provides a polarity estimation system, an information distribution system, a polarity estimation method, and a polarity estimation program that can determine the evaluation polarity of reputation information without registering evaluation expressions for all evaluation expressions in advance. And a means for solving the problems with the purpose of providing a program for estimating the evaluation polarity
- a first polarity estimation system is a polarity estimation system that estimates an evaluation polarity indicating whether reputation information is positive or negative, and includes reputation information with a known evaluation polarity.
- a reputation information storage unit stored in advance, and polarity estimation means for estimating an evaluation polarity of reputation information whose evaluation polarity is unknown based on reputation information stored in advance in the reputation information storage unit. .
- the second polarity estimation system of the present invention is a polarity estimation system that inputs reputation information and estimates an evaluation polarity indicating a force that the input reputation information is positive or negative.
- An evaluation expression storage unit that stores an evaluation polarity corresponding to an evaluation expression that is an expression indicating evaluation of an object
- a reputation information storage unit that stores reputation information and an evaluation polarity corresponding to the reputation information
- the polarity estimation unit estimates the evaluation polarity for the input reputation information. 101).
- the third polarity estimation system of the present invention inputs an object to be evaluated, an attribute expression indicating the attribute of the object, and reputation information including an evaluation expression that is an expression indicating the evaluation of the object.
- estimation means for estimating the evaluation polarity for the inputted reputation information based on the reputation information stored in the evaluation expression storage unit and the reputation information stored in the reputation information storage unit.
- the polarity estimation means calculates a degree of polarity indicating the degree of reputation information positive /! Or the degree of negative! / As the evaluation polarity.
- the fourth polarity estimation system of the present invention has a polarity indicating whether the information to be estimated belongs to! /, The concept of deviation when the information to be estimated can be classified by two predetermined concepts.
- a polarity estimation system for estimation based on an information storage unit that stores information with a known polarity in advance and information with a known polarity stored in advance by the information storage unit! / Polarity estimation means for estimating the polarity of certain information.
- a first information distribution system of the present invention includes a reputation information distribution system that distributes reputation information and an evaluation polarity estimation system that estimates an evaluation polarity that indicates whether the reputation information is positive or negative.
- the evaluation polarity estimation system includes a reputation information storage unit that stores reputation information with a known evaluation polarity in advance, and evaluation information based on reputation information with a known evaluation polarity that the reputation information storage unit stores in advance.
- the reputation information distribution system transmits the evaluation polarity estimated by the evaluation polarity estimation system to the user terminal via the communication network together with the reputation information.
- Information distribution means is provided.
- a first polarity estimation method of the present invention is a polarity estimation method for estimating an evaluation polarity indicating whether reputation information is positive or negative, and the reputation information has a known evaluation polarity. And a polarity estimation step for estimating the evaluation polarity of reputation information whose evaluation polarity is unknown based on the reputation information whose evaluation polarity is previously stored.
- a second polarity estimation method of the present invention is a polarity estimation method for inputting reputation information and estimating an evaluation polarity indicating whether the input reputation information is positive or negative.
- An evaluation expression storing an evaluation polarity corresponding to the evaluation expression which is an expression indicating the evaluation of the object, a reputation information storing step storing the reputation information, and an evaluation polarity corresponding to the reputation information, and storing Based on reputation information for which the evaluation polarity and the stored evaluation polarity are known! /,
- a polarity estimation step for estimating the evaluation polarity with respect to the inputted reputation information; including.
- the third polarity estimation method of the present invention inputs reputation information including an object to be evaluated, an attribute expression indicating the attribute of the object, and an evaluation expression that is an expression indicating the evaluation of the object.
- a polarity estimation step for estimating the evaluation polarity, and in the polarity estimation step the degree of polarity indicating the degree of reputation information positive or negative! / Is calculated as the evaluation polarity.
- a first polarity estimation program of the present invention is a polarity estimation program for estimating an evaluation polarity indicating whether reputation information is positive or negative.
- Reputation information storage process that stores reputation information with known polarity in advance, and polarity estimation that estimates evaluation polarity of reputation information with unknown evaluation polarity based on reputation information with previously stored evaluation polarity Processing.
- the second polarity estimation program of the present invention inputs reputation information, and polarity estimation for estimating an evaluation polarity indicating the power of the input reputation information being positive or negative.
- An evaluation expression storage process for storing an evaluation polarity corresponding to an evaluation expression that is an expression indicating evaluation of an object, reputation information, and an evaluation polarity corresponding to the reputation information are stored in a computer Reputation information storage processing and polarity estimation processing for estimating evaluation polarity with respect to input reputation information based on the stored evaluation polarity and the reputation information for which the stored evaluation polarity is known.
- the third polarity estimation program of the present invention provides reputation information including an object to be evaluated, an attribute expression indicating the attribute of the object, and an evaluation expression that is an expression indicating the evaluation of the object.
- a polarity estimation program for estimating the evaluation polarity indicating whether the input reputation information is positive or negative, and stores the evaluation polarity corresponding to the evaluation expression in the computer.
- Evaluation information storage processing, reputation information, and reputation information storage processing for storing the evaluation polarity corresponding to the reputation information, stored evaluation polarity and stored evaluation
- the polarity estimation process for estimating the evaluation polarity is executed for the entered reputation information, and the degree of reputation information is positive as the evaluation polarity in the polarity estimation process! This is for executing the process of calculating the degree of polarity indicating /, or the degree of negativeness! /.
- the first evaluation polarity estimation program of the present invention includes reputation information including an object to be evaluated, an attribute expression indicating the attribute of the object, and an evaluation expression that is an expression indicating the evaluation of the object.
- An evaluation polarity estimation program installed in a computer that outputs an evaluation polarity indicating whether the input reputation information is positive or negative, and input processing for inputting reputation information to the computer
- a process for calculating the polarity of an attribute expression included in reputation information with a known evaluation polarity a process for calculating the polarity of an object included in reputation information with a known evaluation polarity, and an evaluation polarity.
- a polarity estimation system an information distribution system, a polarity estimation method, and a polarity estimation program that can determine the evaluation polarity of reputation information without registering evaluation expressions for all evaluation expressions in advance. And an evaluation polarity estimation program.
- the evaluation polarity is estimated by calculating the evaluation polarity degree using a statistical method based on the following several hypotheses in reputation information.
- the evaluation polarity is a numerical value indicating power that is positive reputation information or negative reputation information.
- the evaluation polarity is a real number between 1 and 1. In this case, the closer the evaluation polarity is to 1, the more positive it is, and the closer it is to 1, the more negative it is.
- the evaluation polarity is also simply referred to as polarity. Note that this is an example, and the evaluation polarity may be a discrete numerical value instead of a continuous numerical value such as “0 0” to “0”. Also good.
- Evaluation expressions include expressions that can determine the polarity in advance, and reputations that include these expressions
- the polarity of information tends to be the same as the polarity of evaluation expressions. As described above, when the evaluation expression is “large”, the polarity cannot be determined. However, there are cases where the polarity can be determined from the evaluation expression alone. For example, since the evaluation expressions “good” and “great” are clearly positive evaluation expressions, the polarity of reputation information including these evaluation expressions is considered to be positive. On the other hand, since the evaluation expressions “bad” and “dirty” are clearly negative evaluation expressions, similarly, the polarity of reputation information including these evaluation expressions is negative.
- An evaluation polarity estimation system to which the present invention is applied using hypothesis 1 and hypothesis 2 includes an evaluation information storage unit, an evaluation expression storage unit, and polarity estimation means.
- the polarity estimation unit inputs reputation information, refers to the degree of polarity of reputation information stored in the reputation information storage unit, and the evaluation expression and polarity degree stored in the evaluation expression storage unit, and the polarity is unknown. Calculate the polarity of reputation information.
- the polarity estimation means first refers to reputation information whose polarity is known, and sets the degree of polarity of the evaluation expression, the degree of polarity of the attribute expression, and the combination of the attribute expression and the evaluation expression included in the input reputation information. Calculate the degree of polarity.
- the degree of polarity is calculated using the ratio of the number of reputations.
- output a degree of polarity that further combines these calculated degrees of polarity.
- the polarity estimation means is based on reputation information whose polarity is known! /, Attribute expression is good !, used in the image! /, Or bad! /
- the object of the present invention can be achieved by calculating the degree of polarity taking into account whether or not it is used in the image! [0037] Hypothesis 3) If there is enough reputation information! /, The ratio between the number of positive reputation information and the number of negative reputation information for each object calculated only from reputation information with known polarity is , Tend to reflect the overall percentage of reputation information. For example, the evaluation of a specific personal computer can grasp the tendency that there are many positive opinions as an overall trend that is promising. This trend is calculated from reputation information with known polarity. In other words, it is possible to estimate the polarity assuming that reputation information whose polarity is unknown has the same tendency.
- the evaluation polarity estimation system to which the present invention is applied includes a reputation information storage unit, an evaluation expression storage unit, and a polarity estimation unit.
- Input the information refer to the reputation information and the degree of polarity stored in the reputation information storage unit, the evaluation expression and the degree of polarity stored in the evaluation expression storage unit, and calculate the polarity of the reputation information whose polarity is unknown.
- the polarity estimation means refers to reputation information having a known polarity, and the degree of polarity of the evaluation expression, the degree of polarity of the object, and the combination of the object and the evaluation expression included in the input reputation information. Calculate the degree of polarity. Referencing reputation information with known polarity, the number of positive reputation information and the number of negative reputation information for each evaluation expression, each object, and each combination of evaluation expression and object, the ratio Etc. are used to calculate the degree of polarity. Next, the evaluation expression included in the input reputation information, the object, and each degree of polarity calculated earlier are compared, and the degree of polarity is output.
- the object of the present invention can be achieved by adopting the configuration as described above and calculating the degree of polarity by the polarity estimation means based on reputation information whose polarity is known.
- an evaluation polarity estimation system to which the present invention is applied includes a reputation information storage unit, an evaluation expression storage unit, and a polarity estimation unit using the above hypothesis 1, hypothesis 2 and hypothesis 3,
- the polarity estimation means inputs reputation information, refers to the reputation information and degree of polarity stored in the reputation information storage unit, and the evaluation expression and degree of polarity stored in the evaluation expression storage unit. Calculate the polarity of information.
- the polarity estimation means includes a degree of polarity of each evaluation expression, a degree of polarity of each attribute expression, a degree of polarity of a combination of the attribute expression and the evaluation expression, a degree of polarity of a combination of the object and the evaluation expression, Genus Calculate the polarity of the combination of sex expression and evaluation expression. Referencing reputation information with known polarity, for each evaluation expression, for each attribute expression, for each object, for each combination of evaluation expression and attribute expression, for each combination of evaluation expression and object, for evaluation expression, attribute expression, and target For each product combination, the degree of polarity is calculated using the number of positive reputation information, the number of negative reputation information, the ratio, etc. Next, the evaluation expression, attribute expression, and object included in the input reputation information are compared with the previously calculated polarities, and the polarities are output.
- Reputation information may change over time.
- the reputation of the object is expected to change gradually over time. For example, the reputation of a soccer player at a certain time varies depending on the score and contribution to winning or losing in the previous game. Therefore, when estimating the polarity of reputation information, it is necessary to consider the passage of time, such as weighting the polarity in recent reputation information.
- the evaluation polarity estimation system to which the present invention is applied uses the above hypothesis 4, and uses a reputation information storage unit, an evaluation expression storage unit, polarity estimation means, In addition, the polarity estimation means calculates the degree of polarity by adding a weight to the latest evaluation information stored in the reputation information storage unit.
- Evaluation of reputation information may vary depending on the type of evaluator.
- the type of evaluator is the gender, age, address, occupation, hobby, purchase merchandise history, etc. of the evaluator.
- the evaluation of a product may vary depending on the type of such evaluator. For example, it is popular among women, but not popular with men, power S, which is popular with evaluators who purchase several PCs with products and hobbies, and unpopular with other evaluators There are personal computers. Therefore, when estimating the polarity of reputation information also, there must force s consider the type of evaluators.
- An evaluation polarity estimation system to which the present invention is applied includes a reputation information storage unit, an evaluation expression storage unit, and an evaluator type storage using the above hypothesis 5 in addition to the configuration of the evaluation polarity estimation system described above. And a polarity estimation unit, and the polarity estimation unit further calculates the degree of polarity for each type of the evaluator with reference to the evaluator type storage unit.
- the evaluation polarity of reputation information whose evaluation polarity is unknown is estimated based on reputation information whose evaluation polarity is stored in advance! Therefore, the evaluation polarity is unknown.
- the evaluation polarity is estimated by using reputation information whose evaluation polarity is known. Therefore, the evaluation polarity of reputation information can be determined without registering evaluation expressions for all evaluation expressions in advance.
- a predetermined weighting process is performed on the evaluation polarity corresponding to the reputation information to be stored based on the acquisition time information indicating when the reputation information is acquired! If it is configured to estimate the evaluation polarity of reputation information whose evaluation polarity is unknown based on the evaluation polarity that has been weighted, the polarity estimation of reputation information can be performed in consideration of temporal changes in reputation information.
- the evaluation polarity of the reputation information whose evaluation polarity is unknown is estimated. If configured, it is possible to estimate the polarity of reputation information in consideration of the bias due to the evaluator type of reputation information.
- FIG. 1 is a block diagram showing an example of the configuration of the polarity estimation system according to the present invention.
- the polarity estimation system is an evaluation polarity estimation system that estimates the evaluation polarity of reputation information.
- the evaluation polarity estimation system can be applied to, for example, a questionnaire automatic counting system for automatically counting questionnaires and an information service system for distributing reputation information and evaluation polarity.
- the evaluation polarity estimation system includes a data processing device 100 that operates under program control, a storage device 200 that stores information, an input means 300, and an output means 400.
- the evaluation polarity estimation system is specifically realized by an information processing device such as a workstation or a personal computer that operates according to a program.
- the input means 300 is realized by an input device such as a keyboard or a mouse included in the information processing device.
- the input means 300 is operated by, for example, a user when inputting reputation information to be evaluated.
- the input means 300 is a network interface provided in the information processing apparatus. It is realized by the interface part.
- the output unit 400 is realized by a display device such as a display device.
- the output unit 400 has a function of outputting (for example, displaying) an estimation result of evaluation polarity of reputation information.
- the output means 400 may be realized by a network interface unit included in the information processing apparatus when outputting the estimation polarity estimation result via the communication network.
- the output unit 400 may be a printing device such as a printer.
- the data processing device 100 is realized by a CPU of an information processing device that operates according to a program.
- the data processing apparatus 100 includes polarity estimation means 101.
- the storage device 200 is realized by a database device such as a magnetic disk device or an optical disk device.
- Storage device 200 includes an evaluation expression storage unit 201 and a reputation information storage unit 202. Each of these means generally operates as follows.
- the evaluation expression storage unit 201 stores in advance an evaluation expression whose evaluation polarity is known.
- FIG. 2 is an explanatory diagram showing an example of evaluation expressions and evaluation polarities stored in the evaluation expression storage unit 201.
- the evaluation expression storage unit 201 is a database that stores an evaluation expression and a degree of polarity (evaluation polarity) in association with each other.
- the degree of polarity is a value from “1” to “1 1”, and the closer the degree of polarity is to “1”, the more positive the evaluation expression. Also, the closer the degree of polarity is to “1 1”, the more negative the evaluation expression.
- evaluation polarity shown in FIG. 2 is an example, and other polarities such as values of “100” to “0” may be used as the degree of polarity. It is also possible to treat numerical values discretely and indicate the evaluation polarity with a symbol such as ⁇ O '' or ⁇ X ''! /, And the evaluation polarity as an affirmation degree column and a negative degree! / Column. It may be shown separately.
- the reputation information storage unit 202 stores reputation information and the degree of polarity (evaluation polarity) output by the polarity estimation means 101.
- FIG. 3 is an explanatory diagram showing an example of reputation information and evaluation polarity stored in the reputation information storage unit 202.
- the reputation information storage unit 202 is a database that stores the reputation information represented by the triplet of the object, the attribute expression, and the evaluation expression and the degree of polarity of the reputation information in association with each other. Note that the reputation information and the polarity stored in the reputation information storage unit 202 are updated as needed based on the polarity output from the polarity estimation means 101.
- FIG. 3 is an explanatory diagram showing another example of reputation information and evaluation polarity stored in the reputation information storage unit 202. As shown in FIG. 4, the reputation information storage unit 202 may store an affirmative degree and a negative degree instead of the polar degree as the evaluation polarity.
- the polarity estimation means 101 has a function of inputting reputation information and outputting the degree of polarity of the input reputation information.
- FIG. 5 is a block diagram showing an example of the configuration of the polarity estimation means 101. As shown in FIG. As shown in FIG. 5, the polarity estimation means 101 includes a polarity degree reference means 1011, an individual polarity degree calculation means 1012, a total polarity degree calculation means 1013, and a polarity degree registration means 1014.
- the polarity reference means 1011 receives (inputs) reputation information from the input means 300, and searches and determines whether or not the evaluation expression included in the input reputation information is in the evaluation expression storage unit 201. It has a function. If the degree-of-polarity reference means 1011 determines that the reputation information stored in the evaluation expression storage unit 201 matches the evaluation expression included in the input reputation information, the degree-of-polarity reference means 1011 corresponds to the evaluation expression determined to match. The function of extracting the degree of polarity to be extracted from the evaluation expression storage unit 201 is provided. The degree of polarity extracted by the degree-of-polarity reference means 1011 from the evaluation expression storage unit 201 is also referred to as an evaluation expression polarity degree.
- the individual polarity degree calculation means 1012 has a function of inputting reputation information and obtaining the degree of polarity by referring to the reputation information storage unit 202. In this case, the individual polarity degree calculation means 1012 calculates the degree of polarity for each object, attribute expression or evaluation expression. Further, the individual polarity degree calculating means 1012 calculates the degree of polarity for any combination of the object, the attribute expression or the evaluation expression, or two or all combinations.
- the individual polarity calculation means 1012 calculates the polarity of the object as follows.
- the individual polarity degree calculation means 1012 refers to the reputation information storage unit 202, and selects the object whose polarity degree is to be calculated.
- the degree of polarity of all of the included reputation information is extracted from the reputation information storage unit 202.
- the individual polarity calculation means 1012 calculates the polarity of the object by obtaining the average of the extracted polarities.
- the degree of polarity of an object is also obtained when the degree of polarity of an attribute expression or evaluation expression, the power of an object, attribute expression or evaluation expression, or the degree of polarity of two or all combinations is obtained.
- the degree of polarity can be obtained as in the case. That is, the individual polarity degree calculation means 1 012 refers to the reputation information storage unit 202, and the attribute expression or evaluation expression of the degree of polarity calculation object, or any of the object, attribute expression or evaluation expression, two or all The degree of polarity of all the reputation information including the combination of is extracted from the reputation information storage unit 202. Then, the individual polarity degree calculating means 1012 obtains the polarity degree by obtaining the average of each extracted degree of polarity.
- the individual polarity degree calculation means 1012 may obtain the degree of polarity by obtaining the sum of the degree of polarity extracted from the reputation information storage unit 202. .
- the individual polarity degree calculation means 1012 is based on the number of reputation information having a polarity greater than a certain value and the number of reputation information having a polarity less than a certain value! The ratio and probability of reputation information of less than or a certain value may be obtained and used as the degree of polarity.
- the individual polarity calculation means 1012 firstly extracts all pieces of reputation information stored in the reputation information storage unit 202 that match the inputted reputation information of the evaluation object.
- the individual polarity degree calculation means 1012 secondarily selects reputation information whose corresponding degree of polarity is equal to or greater than a predetermined value (for example, 0.3) from the reputation information extracted primarily.
- the individual polarity degree calculation means 1012 obtains the ratio of the number of reputation information selected secondarily (the number of reputation information with positive polarity) to the number of reputation information extracted first.
- the individual polarity degree calculation means 1012 secondarily selects reputation information whose corresponding degree of polarity is equal to or less than a predetermined value (for example, 0.3) from among the reputation information extracted primarily.
- the individual polarity calculation means 1012 obtains the ratio of the number of reputation information selected secondarily (the number of reputation information with negative polarity) to the number of reputation information extracted first.
- the database included in the evaluation polarity estimation system stores it. Even if there is a bias in the information (in this example, the reputation information and the degree of polarity stored in the reputation information storage unit 202), the polarity can be determined more accurately.
- the degree of polarity calculation means 1012 may calculate only the degree of polarity that can be calculated from the two components of reputation information (object, attribute expression or evaluation expression! /, Or two). ! /
- the individual polarity calculation means 1012 includes attribute expressions in the input evaluation information of the evaluation object. Even if it is, the individual polarity for the attribute expression cannot be calculated. Accordingly, in this case, the individual polarity degree calculation means 1012 may obtain only the individual polarity degree for the object or the evaluation expression and the individual polarity degree for the combination of the object and the evaluation expression.
- the total polarity degree calculation means 1013 inputs the polarity degree (evaluation table present polarity degree) extracted by the polarity degree reference means 1011 and the individual polarity degree calculated by the individual polarity degree calculation means 1012 and inputs them. It has a function to calculate the degree of polarity (also called the total polarity) that combines the evaluation expression polarity and individual polarity. In this case, the total polarity calculation means 1013 adds, for example, the average of each polarity degree (each individual polarity degree) calculated by the individual polarity degree calculation means 1012 to the polarity degree extracted by the polarity degree reference means 1011. Calculate the total polarity.
- the total polarity calculation means 1013 calculates the total polarity by calculating the average of the evaluation expression polarity and each individual polarity. You may ask for. Further, the total polarity degree calculation means 1013 may obtain the total polarity degree by, for example, obtaining the sum of the evaluation expression polarity degree and each individual polarity degree. The total polarity calculation means 1013 may obtain the total polarity by performing predetermined weighting on the evaluation expression polarity and each individual polarity.
- the total polarity degree calculation means 1013 gives a large weight to the individual polarity degree for the input reputation information of the evaluation object and all the elements of the object, the attribute expression and the evaluation expression (specifically, The total polarity may be obtained by multiplying a large weighting factor).
- Polarity registration means 1014 includes reputation information to be evaluated and total polarity calculation means 1013. It has a function of associating the calculated degree of polarity (total degree of polarity) with the reputation information storage unit 202 and storing it.
- FIG. 6 is a flowchart illustrating an example of processing in which the evaluation polarity estimation system estimates the evaluation polarity.
- the data processing apparatus 100 of the evaluation polarity estimation system inputs reputation information to be evaluated from the input means 300 in accordance with a user operation (step S10).
- the reputation information is information that is a combination of the object, the attribute expression, and the evaluation expression. For example, information that combines three elements such as reputation information (X computer, noise, dislike) and reputation information (X computer, noise, large) is input.
- the reputation information is expressed by surrounding it with a turtle shell [].
- the three elements separated by punctuation marks represent the object, attribute expression, and evaluation expression, respectively.
- the reputation information may not include any power of the object or the attribute expression.
- the data processing apparatus 100 passes the input reputation information of the evaluation target to the polarity reference means 1011 of the polarity estimation means 101.
- the polarity reference means 1011 refers to the evaluation expression storage unit 201 and acquires (extracts) the polarity degree of the evaluation expression included in the reputation information of the evaluation target from the evaluation expression storage unit 201 (step) Sll).
- the evaluation expression storage unit 201 stores the evaluation expression and the polarity shown in FIG. 2
- the degree-of-polarity reference means 1011 receives (inputs) the reputation information of the evaluation target [X X computer, noise, dislike]
- the evaluation expression storage unit 201 is referred to and the evaluation expression “dislike! /,” Acquire (extract) the polarity degree “1” corresponding to.
- the evaluation expression storage unit 201 stores the evaluation expression stored in the evaluation expression. Since the evaluation expression “large” does not exist, the degree of polarity is set to “0”. The degree of polarity “0” means that the evaluation polarity is unknown.
- the polarity estimation unit 101 holds the polarity degree extracted by the polarity degree reference unit 1011 in a storage unit such as a memory, and passes the reputation information of the evaluation target input from the input unit 300 to the individual polarity degree calculation unit 1012 .
- the individual polarity degree calculation means 1012 receives (inputs) the reputation information to be evaluated, refers to the reputation information storage unit 202, and associates the reputation information and the degree of polarity with the reputation information storage unit 2. All are acquired (extracted) from 02 (step S12).
- the individual polarity degree calculation means 1012 receives (inputs) reputation information [X computer, noise, large], the reputation information storage unit 202 is referred to, and the object “X computer”, attribute expression “ Reputation information including “noise” or evaluation expression “large V,” and the corresponding degree of polarity are acquired (extracted) from the reputation information storage unit 202.
- the reputation information storage unit 202 stores the reputation information and polarity shown in FIG. 3
- the individual polarity calculation means 1012 includes the object, attribute expression of the first record, the fifth record, and the sixth record, Obtain (extract) evaluation expression and degree of polarity.
- the individual polarity calculation means 1012 receives the reputation information of the evaluation target input in step S10 (hereinafter also referred to as input reputation information), the reputation information acquired (extracted) in step S12, and the reputation information. Based on the degree of polarity corresponding to the reputation information of the object, the attribute expression or evaluation expression, or the power of the object, attribute expression or evaluation expression, or the polarity degree for two or all combinations, Eventually, one or more forces are calculated (step S13).
- One of the degrees of polarity for “ ⁇ X PC—large”, the combination of attribute expression and evaluation expression “noise—large”, and the object, attribute expression and evaluation expression group “ ⁇ X PC noise is greatest” Calculate one or more.
- the individual polarity degree calculation means 1012 calculates the degree of polarity for the object, the degree of polarity for the attribute expression, and the degree of polarity for the evaluation expression.
- the individual polarity degree calculation means 1012 assigns “ ⁇ X PC” to the object out of the degree of polarity for the reputation information acquired (extracted) in step S12.
- the individual polarity is calculated by calculating the average polarity for the reputation information included.
- the individual polarity degree calculation means 1012 calculates the polarity degree (individual polarity degree) of the object using the equation (1).
- Np represents the number of reputation information including the object
- Pi represents the degree of polarity of each piece of reputation information including the object.
- the individual polarity degree calculation means 1012 uses the object “O X PC” and the evaluation expression “Large”. The individual polarity is calculated by calculating the average of the polarities for the reputation information included. Similarly, the individual degree-of-polarity calculation means 1012 obtains the degree of polarity with respect to the combination of the attribute expression and the evaluation expression “noise—large” and the object, the attribute expression and the evaluation expression “O X PC—noise—high”. In this case, the individual polarities are calculated by calculating the average of the polarities for the reputation information including all of the object “X PC”, attribute expression “noise”, and evaluation expression “large”.
- the individual polarity degree calculation method described above is an example, and the individual polarity degree calculation means 1012, for example, obtains the sum of the polarity degrees extracted from the reputation information storage unit 202.
- the individual polarity degree may be obtained.
- the individual polarity degree calculation means 1012 is based on the number of reputation information with a polarity greater than a certain value and the number of reputation information with a polarity less than a certain value. The ratio or probability of information or reputation information below a certain value may be obtained and used as the degree of polarity.
- the individual polarity calculation means 1012 may only calculate the degree of polarity that can be calculated from the two components of reputation information (object, attribute expression or evaluation expression !, or two of them).
- the individual polarity degree calculation means 1012 obtains the individual polarities of all of these objects, attribute expressions or evaluation expressions, or any of these objects, attribute expressions or evaluation expressions, two or all combinations. There is no need to calculate.
- the individual polarities include the polarities for the object, the polarities for the attribute expression, the polarities for the evaluation expression, the polarities for the combination of the object and the attribute expression, and the combinations of the object and the evaluation expression.
- the individual polarity degree calculation means 1012 for example, only has to calculate three degrees: the polarity degree for the object, the polarity degree for the attribute expression, and the polarity degree for the evaluation expression! /.
- the individual polarity degree calculation means 1012 passes the calculated individual polarity degree to the total polarity degree calculation means 1013.
- the total polarity calculation means 1013 inputs the polarity (evaluation expression polarity) acquired (extracted) in step SI 1 and the individual polarity calculated in step S13, and these evaluation expression polarities are input.
- the degree of polarity (total degree of polarity) that combines the degree and the individual degree of polarity is calculated (step S14). For example, when calculating the integrated polarity (total polarity), the total polarity calculation means 1013 adds the value obtained by averaging the individual polarities calculated in step S12 to the polarity acquired in step S11. To do.
- the polarity power obtained in step S11 is “0”.
- the polarities for the object are “10.3”
- the polarities for the attribute expression are “10.8”
- the polarities for the evaluation expression are “10.8”.
- the total polarity calculation means 1013 calculates the total polarity (total polarity) as “1. 3”.
- the calculation method as described above is used based on the idea that the degree of polarity of the evaluation expression is corrected by the individual polarity degree.
- the calculation method of the total polarity shown is an example, and the total polarity may be obtained simply by calculating the average of the evaluation expression polarity and the individual polarity or by calculating the total.
- the polarity degree registration means 1014 additionally registers the input reputation information input in step S10 and the polarity degree (total polarity degree) calculated in step S14 in the reputation information storage unit 202 (step S 15).
- the polarity degree registration means 1014 stores the reputation information and the polarity degree in the reputation information storage unit 202 in association with each other. For example, if the reputation information is [X PC, noise, large] and the degree of polarity is “one 0.3”, the polarity degree registration means 1014 newly adds a record having these elements as elements.
- the polarity estimation means 101 causes the output means 400 to output the degree of polarity (step S16).
- the polarity estimation means 101 may output a numerical value such as “1.0.3” as the degree of polarity, and if the degree of polarity is a value equal to or greater than a certain threshold, The symbol “X” or the like may be output if it is below. Furthermore, the individual polarity calculated in step S13 may be output.
- the output means 400 outputs (for example, displays) the degree of polarity in accordance with the instruction from the polarity estimation means 101.
- objects, attribute expressions or evaluation expressions included in reputation information whose evaluation polarity is known, and combinations thereof.
- Each evaluation polarity is calculated for each combination.
- the evaluation polarity is output against the object, attribute expression, and evaluation expression included in the reputation information whose evaluation polarity is unknown. Therefore, the evaluation polarity can be estimated by using the reputation information with the known evaluation polarity for the reputation information with the unknown evaluation polarity.
- the polarity estimation means 101 uses the representation of a good image or a bad image in the attribute expression based on reputation information whose polarity is known only by the polarity of the evaluation expression, Evaluation polarity can be estimated by taking into account the degree of affirmation or denial by the object. Therefore, it is possible to estimate the polarity for an evaluation expression whose evaluation polarity is unknown. In other words, based on the reputation information accumulated so far, the evaluation polarity can be estimated by taking into account the bias of the polarity of the object, attribute expression, and evaluation expression, and the evaluation polarity cannot be determined. Reduce the number of incidents.
- the polarity estimation unit 101 sequentially stores the calculated calculation result of the degree of polarity in the reputation information storage unit 202. Then, the polarity estimation means 101 uses the result of the polarity stored in the reputation information storage unit 202 for the subsequent calculation of the polarity. Therefore, although the accuracy of the degree of polarity calculation is somewhat poor at the start of operation of this system, the results of repeated calculation of the degree of polarity are accumulated, and as the reputation information to be accumulated increases, The accuracy of sex estimation can be improved.
- FIG. 7 is a block diagram illustrating a configuration example of the polarity estimation system (evaluation polarity estimation system) in the second embodiment.
- the information content stored in the reputation information storage unit 203 is different from the information content stored in the reputation information storage unit 202 shown in the first embodiment.
- the function of the polarity estimation means 102 is different from the function of the polarity estimation means 101 shown in the first embodiment.
- the functions of the constituent elements other than the polarity estimation means 102 and the reputation information storage unit 203 are the same as those functions described in the first embodiment.
- the reputation information storage unit 203 stores reputation information, an acquisition date of the reputation information, and a degree of polarity (evaluation polarity) with respect to the reputation information.
- FIG. 8 is an explanatory diagram showing an example of evaluation information, acquisition date, and evaluation polarity stored in the reputation information storage unit 203.
- the reputation information storage unit 203 is a database that stores the time when reputation information is acquired (acquisition date in this example), the object, the attribute expression, the evaluation expression, and the degree of polarity as one record. That is, in the present embodiment, the reputation information storage unit 203 associates reputation information (including the object, attribute expression, and evaluation expression), the acquisition date of the reputation information, and the evaluation polarity for the reputation information.
- the date of acquisition of reputation information is obtained based on a time signal output from a timer provided in the data processing device 100 when the reputation information is registered in the reputation information storage unit 203, for example.
- the device 100 is associated with the reputation information and stored in the reputation information storage unit 20d.
- the degree of polarity is a value from “1” to “1”, and the closer the degree of polarity is to “1”, the more positive the evaluation expression. In addition, the closer the degree of polarity is to “1 1”, the more negative the expression.
- time indicates a date.
- the reputation information and evaluation polarity shown in Fig. 8 are merely examples, and the reputation information and evaluation of the object. You may use a combination of a valence expression and a combination of an attribute expression and an evaluation expression. Also, numerical values may be handled discretely, and the evaluation polarity may be indicated by a symbol such as “ ⁇ ” or “X”, or the evaluation polarity may be divided into a positive degree column and a negative degree column. Motole.
- the time indicating the acquisition date of reputation information may be information other than the date. For example, it may include the time until reputation information is acquired, or information including only the year and month.
- polarity estimation means 102 inputs reputation information to be evaluated, and calculates the degree of polarity obtained by adding a weight to the degree of polarity corresponding to the latest reputation information among the pieces of reputation information accumulated in advance. However, it differs from the first embodiment in that it is output.
- FIG. 9 is a block diagram showing a configuration example of the polarity estimation means 102 in the second exemplary embodiment.
- the polarity estimation means 102 includes the weighting means 1021 in addition to the components of the polarity estimation means 101 shown in FIG. And different.
- the weighting means 1021 inputs the reputation information to be evaluated, refers to the reputation information storage unit 203, and sets the related reputation information, time (date of acquisition of reputation information), and polarity to the reputation information storage unit 203.
- the function to acquire (extract) from is provided.
- the weighting means 1021 extracts all of the reputation information that matches the elements (objects, attribute expressions, evaluation expressions) included in the reputation information of the evaluation target from the reputation information storage unit 203, and adds them to the extracted reputation information. Extract the corresponding time (acquisition date) and degree of polarity.
- the weighting means 1021 calculates the degree of polarity (with both weighted polarities! /, U) with a large weight on the latest reputation information among the extracted reputation information, and the reputation information and weighted polarity.
- the function of passing the degree to the individual polarity degree calculation means 1012 is provided.
- the weighting means 1021 selects, based on the extracted time (acquisition date), reputation information whose acquisition date is within a predetermined number of days from the current date among the extracted reputation information.
- the weighting means 1021 then weights the degree of polarity corresponding to the selected reputation information! / (For example, multiplying by a predetermined weighting factor), and obtains the weighted degree of polarity using the weighted degree of polarity.
- FIG. 10 is a flowchart illustrating a processing example in which the evaluation polarity estimation system according to the second embodiment estimates the evaluation polarity.
- weighting processing step S 17 This is different from the first embodiment in that is added.
- the data processing device 100 of the evaluation polarity estimation system inputs reputation information to be evaluated from the input means 300 in accordance with a user operation (step S10).
- Data processing device 100 of the evaluation polarity estimation system inputs reputation information to be evaluated from the input means 300 in accordance with a user operation (step S10).
- the polarity degree reference means 1011 refers to the evaluation expression storage unit 201 and acquires (extracts) the polarity degree of the evaluation expression included in the input reputation information from the evaluation expression storage unit 201 (step S 11). .
- the polarity estimation unit 102 holds the polarity degree extracted by the polarity degree reference unit 1011 in a storage unit such as a memory, and passes the reputation information of the evaluation target input from the input unit 300 to the weighting unit 1021.
- the weighting means 1021 receives (inputs) the input reputation information input in step S10, refers to the reputation information storage unit 203, and related reputation information, time (reputation information acquisition date) And the degree of polarity are all acquired (extracted) from the reputation information storage unit 203 (step S12).
- the weighting means 1021 receives input reputation information [X computer, noise, large] (when input), it refers to the reputation information storage unit 203, and the object “O X computer”, attribute expression “noise” ”Or 8 records of reputation information including the evaluation expression“ Large ”, the object, attribute expression, evaluation expression, time and polarity are all acquired from the reputation information storage unit 20 3 (extraction). To do.
- the weighting means 1021 calculates the degree of polarity by assigning a large weight to the latest reputation information among the extracted reputation information (step S17). For example, the weighting means 1021 applies a weight of 1 to the degree of polarity for reputation information acquired in a certain period (the last three months, etc.), and applies a weight of 0 to other polarities. For example, if the target is a personal computer, the model is changed every quarter, so only the reputation information evaluated within the latest three months is used to determine the degree of specificity. This is an example.For example, the weight may be changed every month, or the difference between the current time and the time when reputation information is acquired is calculated, and the reciprocal of the obtained time difference is used as the weight coefficient. As well as the degree of polarity. Then, the weighting means 1021 passes the reputation information to be evaluated and the obtained weighted polarity degree to the individual polarity degree calculating means 1012.
- the individual polarity calculation means 1012 is based on the input reputation information input in step S10, the reputation information extracted in step S17, and the calculated weighted polarity.
- the object, attribute expression or evaluation expression, or the force of either the object, attribute expression or evaluation expression, or the degree of polarity for two or all combinations is calculated (step S13).
- the total polarity calculation means 1013 inputs the polarity (evaluation expression polarity) acquired (extracted) in step SI 1 and the individual polarity calculated in step S13, and these evaluation expression polarities are input.
- the degree of polarity (total degree of polarity) that combines the degree and the individual degree of polarity is calculated (step S14).
- the polarity registration means 1014 stores the input reputation information input in step S10, the polarity (total polarity) calculated in step S14, and the current time in the reputation information storage unit 203. Additional registration is performed (step S15).
- the polarity degree registration means 1014 stores the reputation information, the polarity degree, and the current time in the reputation information storage unit 203 in association with each other.
- the polarity estimation means 101 causes the output means 400 to output the degree of polarity (step S16).
- the above-described configuration for weighting is an example, and for example, a configuration in which the individual polarity degree calculation unit 1012 has a function substantially similar to that of the weighting unit may be used. . In other words, the configuration for weighting is not limited to the one shown above.
- the weighting means 1021 calculates the degree of polarity by assigning a greater weight to the degree of polarity of the latest reputation information. Therefore, in addition to the effects shown in the first embodiment, the power S can be used to estimate the polarity of reputation information in consideration of temporal changes in reputation information.
- FIG. 11 is a block diagram showing a configuration example of a polarity estimation system (evaluation polarity estimation system) in the third embodiment.
- the information content stored in the reputation information storage unit 204 is stored in the reputation information storage unit 202 shown in the first embodiment. Different from information content.
- the function of the polarity estimation means 103 is different from the function of the polarity estimation means 101 shown in the first embodiment.
- the present embodiment is different from the first embodiment in that the storage device 200 includes an evaluator type storage unit 205 in addition to the components shown in FIG.
- the functions of the constituent elements other than the polarity estimation means 103, the reputation information storage unit 204, and the evaluator type storage unit 205 are the same as those functions described in the first embodiment.
- the reputation information storage unit 204 stores reputation information, an evaluator ID for identifying an evaluator who evaluated the reputation information, and a polarity (evaluation polarity) for the reputation information.
- FIG. 12 is an explanatory diagram showing an example of reputation information, evaluator ID, and evaluation polarity stored in the reputation information storage unit 204.
- the reputation information storage unit 204 is a database that includes an evaluator ID, an object, an attribute expression, an evaluation expression, and a degree of polarity of an evaluator who has entered an evaluation of reputation information in one record.
- the reputation information storage unit 204 includes reputation information (including the object, attribute expression, and evaluation expression), the evaluator ID of the evaluator who evaluated the reputation information, and the evaluation for the reputation information.
- the polarity is stored in association with each other.
- the evaluator ID is stored in the reputation information storage unit 204 in association with the reputation information by the data processing device 100 when the reputation information is registered in the reputation information storage unit 204.
- the degree of polarity is a value from “1” to “1”, and the closer the degree of polarity is to “1”, the more positive the evaluation expression. Also, the closer the degree of polarity is to “1 1”, the more negative the evaluation expression.
- the evaluator ID stored in the reputation information storage unit 204 corresponds to the evaluator ID stored in the evaluator type storage unit 205 described later!
- the reputation information and evaluation polarity shown in Fig. 12 are examples, and as reputation information, there are two sets of object and evaluation expression, and two sets of attribute expression and evaluation expression. May be used. Also, numerical values may be handled discretely, and the evaluation polarity may be indicated by a symbol such as “ ⁇ ” or “X”, or the evaluation polarity may be divided into a positive degree column and a negative degree column. As described above, the degree of polarity may be expressed in other ways.
- the evaluator type storage unit 205 stores evaluator type information that is information indicating the type of the evaluator.
- FIG. 13 is an explanatory diagram showing an example of the evaluator type information stored in the evaluator type storage unit 205.
- the evaluator type storage unit 205 is a database that includes an evaluator ID and the gender, age, occupation, and hobby of the evaluator of the evaluator ID in one record. That is, in the present embodiment, the evaluator type storage unit 205 stores gender, age, occupation, and hobby as the evaluator type in association with the evaluator ID of the evaluator.
- a blank portion indicates that the type is unknown.
- the hobbies are delimited by “,”, indicating that the evaluator type storage unit 205 can store a plurality of hobbies for the evaluator.
- the evaluator type information shown in FIG. 13 is an example, and the evaluator type storage unit 205 may store other information such as a purchase product history as the evaluator type information.
- the polarity estimation means 103 inputs the reputation information to be evaluated and the evaluator type of the evaluator who evaluated the reputation information, and adds the functions shown in the first embodiment. In addition, it has a function to calculate the degree of polarity for each evaluator type and output the degree of polarity considering the bias for each evaluator type.
- FIG. 14 is a block diagram illustrating a configuration example of the polarity estimation unit 103 according to the third embodiment.
- the polarity estimation means 103 includes the type polarity degree calculation means 1031 in addition to the components of the polarity estimation means 101 shown in FIG. The form is different.
- the order of the type polarity degree calculation means 1031 and the individual polarity degree calculation means 1012 may be reversed.
- the type polarity calculation means 1031 inputs the evaluator type and reputation information, refers to the evaluator type storage unit 205 and the reputation information storage unit 204, and sets the evaluator type such as age and gender. It has a function to calculate the degree of polarity for each combination with reputation information (hereinafter referred to as evaluator type polarity degree! / ⁇ ⁇ ). For example, if the evaluator type is gender, age, occupation, hobbies, and purchased product history, the type polarity degree calculation means 1031 includes a combination of the target and gender, a combination of the target and age, and a target and occupation.
- the degree of polarity (evaluator type polarity) is calculated for the combination of the object, the combination of the object and the hobby, and the combination of the object and the purchased product name. By doing so, for the entered evaluation expression, a rating of a similar evaluator type is given. It is possible to calculate what kind of evaluation the price person is doing.
- the type polarity degree calculation means 1031 first determines which combination degree of polarity is to be calculated.
- the degree of polarity for the combination of gender and object and the combination of hobby and object shall be calculated.
- the type polarity degree calculation means 1031 may determine which combination degree of polarity is to be determined according to, for example, a user input operation, and based on preset setting information. You can decide!
- the type polarity degree calculation means 1031 refers to the evaluator type storage unit 205 and the reputation information storage unit 204, and the gender is “m” and the object is “X PC”. Acquire (extract) all reputation information and the degree of polarity corresponding to the reputation information. Then, the type polarity degree calculation means 1031 obtains the average of each extracted degree of polarity. Similarly, the type polarity degree calculation means 10 31 obtains (extracts) all the reputation information whose hobby is “PC” and whose target is “X PC” and the degree of polarity corresponding to the reputation information. ) Then, the type polarity degree calculation unit 1031 obtains the average of each extracted degree of polarity.
- the polarity degree calculation method shown above is an example, and the type polarity degree calculation means 1031 is the degree of polarity for the combination of the evaluator type and other elements of the reputation information shown in the present embodiment. You may ask for. Further, the type polarity degree calculation means 1031 may obtain the degree of polarity by obtaining the total instead of the average of the extracted degrees of polarity.
- FIG. 15 is a flowchart illustrating a processing example in which the evaluation polarity estimation system according to the third embodiment estimates the evaluation polarity.
- the present embodiment differs from the first embodiment in that a type polarity degree calculation process (step S18) is added in addition to the processes shown in FIG. .
- step S18 the type polarity degree calculation process
- step S13 the individual polarity degree calculation process
- the data processing device 100 of the evaluation polarity estimation system according to the user's operation, The reputation information to be evaluated and the evaluator type are input from the input means 300 (step S10).
- the data processing apparatus 100 passes the evaluator type information such as the input evaluator ID, gender, age, occupation, hobby, or purchased product history to the type polarity degree calculation means 1031 of the polarity estimation means 103. If the evaluator type storage unit 205 stores evaluator type information in advance, the data processing apparatus 100 passes only the evaluator ID to the type polarity degree calculation means 1031. If the evaluator type information is not stored in advance, the data processing apparatus 100 inputs the evaluator type information and passes it to the type polarity degree calculation means 1031.
- evaluator type information for example, when reputation information is extracted based on a free description questionnaire, an evaluator type item is included in the questionnaire item.
- the collected result power of the questionnaire may also be extracted from the evaluator type information.
- evaluator-type information is obtained using an existing method that determines the gender of the article writer based on how the blog article is written. May be.
- the data processing apparatus 100 passes the input reputation information and the evaluator type to the polarity reference means 1011 of the polarity estimation means 103.
- the polarity degree reference means 1011 refers to the evaluation expression storage unit 201 and acquires (extracts) the degree of polarity of the evaluation expression included in the reputation information from the evaluation expression storage unit 201 (step SI 1).
- the polarity estimation unit 103 holds the polarity, reputation information, and evaluator type extracted by the polarity reference unit 1011 in a storage unit such as a memory.
- the polarity estimation means 103 receives (inputs) the input reputation information and the input evaluator type input at step S10, refers to the reputation information storage unit 204 and the evaluator type storage unit 205, and All the related reputation information, evaluator type, and polarity are acquired (extracted) from the reputation information storage unit 204 and the evaluator type storage unit 205 (step S12).
- the polarity estimation means 103 receives input reputation information [X PC, noise, large] and receives gender “male” and hobby “computer” as input evaluator types (when input), Referring to the reputation information storage unit 204 and the evaluator type storage unit 205, the target object “X X PC”, attribute expression “noise”, evaluation expression “large”, gender “male”, or hobby “computer” Including all the reputation information included (extracted).
- the acquired data includes the object, attribute expression, and rating. This record includes value expression, gender, hobbies and degree of polarity. Then, the polarity estimation unit 103 passes the acquired record to the type polarity degree calculation unit 1031.
- the type polarity degree calculation means 1031 calculates the degree of polarity (evaluator type polarity degree) for each combination of the evaluator type such as age and gender and reputation information (step S 18).
- the type polarity degree calculation means 1031 receives (inputs) the input reputation information and the input evaluator type in step S10 and the record acquired in step S12, and the polarity degree for the combination of age and object, The degree of polarity or the like for the combination of the hobby and the object is calculated.
- the type polarity degree calculation means 1031 first determines which combination degree of polarity is to be calculated.
- the degree of polarity is calculated for the combination of gender and object, and the combination of hobby and object.
- the type polarity degree calculation means 1031 obtains the reputation information that has the gender power S "male” and the object is "X PC" among the records acquired in step S12, and the reputation information. Acquire (extract) all polarities corresponding to. Then, the type polarity degree calculation means 1031 obtains the average of the extracted degrees of polarity. Similarly, the type polarity degree calculation means 1031 acquires (extracts) all the reputation information whose hobby is “PC” and whose target is “X PC” and the degree of polarity corresponding to the reputation information. . Then, the type polarity degree calculation means 1031 obtains the average of the extracted degrees of polarity.
- the polarity degree calculation method shown above is an example, and the type polarity degree calculation means 1031 is the degree of polarity for the combination of the evaluator type and other elements of the reputation information shown in the present embodiment. May be calculated. Further, the type polarity degree calculating means 1031 may obtain the degree of polarity by obtaining the total instead of the average of the extracted degrees of polarity.
- the individual polarity calculation means 1012 receives (inputs) the input reputation information input in step S10 and the record acquired in step S12, and the object, attribute expression or evaluation expression, Alternatively, the power of any one of the object, the attribute expression, and the evaluation expression, or the degree of polarity for two or all combinations is calculated (step S13).
- the total polarity calculation means 1013 obtains (extracts) the polarity (evaluation) obtained in step SI 1. Input the degree of polarity (expression polarity), the degree of polarity (evaluator type polarity) for the combination of the evaluator type and reputation information calculated in step SI8, and the individual degree of polarity calculated in step S13.
- the degree of polarity (total degree of polarity) is calculated by combining the degree of polarity, the evaluator type polarity degree, and the individual degree of polarity (step S14).
- the total polarity calculation means 1013 is integrated by adding the average of the polarities calculated in step S18 and the average of the individual polarities calculated in step S13 to the polarities acquired in step S11. Calculate the degree of polarity (total polarity).
- the total degree of polarity calculation unit 1013 may determine the total degree of polarity by calculating the sum or average of each degree of polarity.
- the polarity degree registration means 1014 stores the input reputation information and the input evaluator type input at step SI 1 and the polarity degree calculated at step S 14 and stores the reputation information storage unit 204 and the evaluator type. It is additionally registered in the part 205 (step S15). In this case, the polarity registration means 1014 stores the reputation information, the polarity, and the evaluator ID in association with each other in the reputation information storage unit 205.
- the polarity estimation means 103 causes the output means 400 to output the degree of polarity (step S16).
- the type polarity degree calculation means 1031 calculates the evaluation tendency by calculating the evaluation tendency for each type of the evaluator. Therefore, in addition to the effects shown in the first embodiment, it is possible to fidelity estimation of reputation information taking into account the bias by reputation information evaluator type.
- FIG. 16 is a block diagram showing a specific configuration example of each evaluation polarity estimation system shown in each of the above embodiments.
- the evaluation polarity estimation system includes a data processing device 100A, a storage device 200A, an input device 300A, an output device 400A, and a program storage device 600.
- data processing apparatus 100 is realized by a computer that operates according to a program.
- the data processing device 100A includes an input device 300A such as a keyboard and a mouse, and a display.
- An output device 400A such as a ray device or a printer is connected.
- a storage device 200A is connected to the data processing device 100.
- the storage device 200A is a device including the evaluation expression storage unit 201, the reputation information storage unit 202, etc., and is connected to the data processing device 100A via a bus or the like! /, May! /, And a communication network. Connected through! /, Even! /
- the storage device 200A also includes an evaluator type storage unit 205.
- the data processing device 100 includes a program storage device (for example, a hard disk device or a CD-ROM) 600 that stores the evaluation polarity estimation program 500.
- a program storage device 600 for example, a hard disk device or a CD-ROM
- the program storage device 600 is based on reputation information storage processing for storing reputation information with a known evaluation polarity in a computer and reputation information with a known evaluation polarity stored in advance.
- a polarity estimation program for executing the polarity estimation processing for estimating the evaluation polarity of reputation information whose evaluation polarity is unknown is stored.
- the data processing device 100A reads the evaluation polarity estimation program 500 from the program storage device 600 and operates according to the read evaluation polarity estimation program 500. By operating in this manner, the data processing apparatus 100A operates as the polarity estimation unit 101, the polarity estimation unit 102, and the polarity estimation unit 103.
- the data processing apparatus 100A which is a computer, may include a storage device inside and store information (for example, input reputation information) in the storage device! /.
- each means evaluation polarity estimation means 101, polarity degree reference means 1011, individual polarity degree calculation means 1012, total polarity degree calculation means 1013, polarity degree registration means 1014,
- the data processing apparatus 100 includes the weighting means 1021 and the type polarity degree calculating means 1031) as separate hardware devices.
- a keyboard and a mouse are shown as examples of the input unit 100.
- reputation information is input from another device to the evaluation polarity estimation system via a communication network.
- a communication interface unit for performing communication via a communication network may be used as the input means 100.
- the output mode of the polarity degree to be output may be a mode in which the polarity degree is output to other devices via the communication network. Yes.
- a communication interface unit for performing communication via a communication network may be used as the output means 400! /.
- the input means 300 is realized by the input device 300A.
- the output means 400 is realized by the output device 400A.
- FIG. 17 is a block diagram showing a configuration example of an information service system according to the present invention.
- the information service system according to the present embodiment includes an evaluation polarity estimation system 1000, a reputation information extraction system 2000, a reputation information service system 3000, an evaluation polarity guesser terminal 400, and a service user terminal 5000.
- the evaluation polarity estimation system 1000, the reputation information extraction system 2000, the reputation information service system 3000, the evaluation polarity predictor terminal 4000, and the service user terminal 5000 are connected via a communication network such as the Internet, for example.
- the evaluation polarity estimation system 1000 is operated by, for example, a service provider (hereinafter also referred to as a reputation information service operator) that provides a reputation information distribution service.
- a service provider hereinafter also referred to as a reputation information service operator
- the evaluation polarity estimation system 1000 is realized by an information processing apparatus such as a workstation or a personal computer that operates according to a program.
- the evaluation polarity estimation system 1000 is any evaluation polarity estimation system from the first embodiment to the third embodiment.
- FIG. 18 is a block diagram showing a configuration example of the polarity estimation system in the fourth exemplary embodiment.
- the evaluation polarity estimation system shown in the first embodiment is applied to an information service system.
- the evaluation polarity estimation system includes reputation information reading means 111 and reputation information writing means 112 in addition to the components shown in the first embodiment.
- the configuration is slightly different from the configuration shown in the first embodiment.
- force S showing the configuration when the evaluation polarity estimation system shown in the first embodiment is applied to the information service system, the second embodiment or The same applies to the case of using the evaluation polarity estimation system shown in the third embodiment.
- the reputation information reading unit 111 and the reputation information writing unit 112 are realized by a CPU and a network interface unit of an information processing apparatus that implements an evaluation polarity estimation system 1000 that operates according to a program.
- the reputation information reading means 11 1 inputs (receives) an object, an attribute expression, and an evaluation expression (reputation information) via a communication network, and receives a reputation information storage unit (reputation information storage unit 202) in the evaluation polarity estimation system 1000. It has a function to read force information.
- the reputation information writing means 112 inputs (receives) an object, attribute expression, evaluation expression, and polarity via a communication network, and inputs the received information to the reputation information storage unit (reputation information) in the evaluation polarity estimation system 1000.
- a function of writing in the storage unit 202) is provided.
- the reputation information extraction system 2000 is operated by, for example, a reputation information service operator, and is specifically realized by an information processing apparatus such as a workstation or a personal computer that operates according to a program.
- the reputation information extraction system 2000 has a function of inputting (receiving) a natural text through a communication network, and extracting and outputting the reputation information.
- the reputation information extraction system 2000 is realized using the existing system described above.
- the reputation information extraction system 2000 includes a database that accumulates reputation information, and extracts reputation information from the database based on the input natural sentence text. Then, the reputation information extraction system 2000 outputs (transmits) the extracted reputation information to the evaluation information service system 3000 via the communication network.
- the reputation information service system 3000 is operated by, for example, a reputation information service operator, and is specifically realized by an information processing apparatus such as a workstation or a personal computer that operates according to a program.
- the reputation information service system 3000 has a function of inputting (receiving) a natural text from a service user terminal 5000 of a service user via a communication network. In addition, the reputation information service system 3000 uses the input natural sentence text to enter the reputation information.
- the information extraction system 2000 has a function to output reputation information. For example, the reputation information service system 3000 outputs (transmits) a natural text to the reputation information extraction system 2000 via a communication network. Then, the reputation information service system 3000 inputs (receives) the reputation information extracted by the reputation information extraction system 2000 from the reputation information extraction system 2000 via the communication network.
- the reputation information service system 3000 has a function of outputting (transmitting) reputation information to the evaluation polarity estimation system 1000 and causing the evaluation polarity estimation system 1000 to output a degree of polarity (evaluation polarity).
- reputation information and evaluation polarity are stored in the evaluation information storage unit (reputation information storage unit 202) in the evaluation polarity estimation system 1000.
- the reputation information service system 3000 has a function of transmitting reputation information and the degree of polarity estimated by the evaluation polarity estimation system 1000 to the service user terminal 5000 via the communication network and presenting it to the service user. Prepare.
- the reputation information service system 3000 sends the reputation information and the degree of polarity in the evaluation polarity estimation system 1000 via the communication network in response to a request from the evaluation polarity thruster terminal 4000 of the evaluation polarity thruster. And output (send) it to the evaluation polarity guesser terminal 4000 and present it, and has the function of prompting the evaluation polarity performer to correct the reputation information and its evaluation polarity.
- the reputation information service system 3000 has a function of recording the amount of money (service usage fee) that the reputation information service operator should receive from the service user and the amount of money (reduction fee) to be paid to the evaluation polarity investigator.
- the reputation information service system 3000 transmits / receives information to / from the service user terminal (service user terminal 5000) and the evaluation polarity investigator terminal (evaluation polarity thruster terminal 4000). Will be described.
- Service user terminal 5000 is a terminal operated by a service user, and is specifically realized by an information processing terminal such as a personal computer.
- the force information service system showing one service user terminal 5000 may include a plurality of service user terminals 5000.
- the service user terminal 5000 may be a mobile terminal such as a mobile phone or a PDA.
- the evaluation polarity thruster terminal 4000 is a terminal operated by the evaluation polarity thruster. Specifically, it is realized by an information processing terminal such as a personal computer.
- the force information service system showing one evaluation polarity thruster terminal 4000 may include a plurality of evaluation polarity thruster terminals 4000.
- the evaluation polarity guesser terminal 4000 may be a mobile terminal such as a mobile phone or a PDA.
- the reputation information service system 3000 includes a control unit 3001 and monetary information storage means 3002.
- the control unit 3001 operates according to a program stored in a storage device (not shown) included in the reputation information service system 3000.
- the control unit 3001 has a function of transmitting / receiving information to / from the service user terminal 5000, the evaluation polarity guesser terminal 4000, the evaluation polarity estimation system 1000, and the reputation information extraction system 2000 via a communication network.
- the reputation information service system 3000 transmits and receives information when communicating with the service user terminal 5000, the evaluation polarity guesser terminal 4000, the reputation information extraction system 2000, and the evaluation polarity estimation system 1000.
- the illustration of the communication interface unit is omitted. Therefore, specifically, the control unit 3001 transmits / receives information to / from other devices via a communication interface unit (not shown).
- the monetary information storage means 3002 is specifically realized by a database device such as a magnetic disk device or an optical disk device.
- the monetary information storage means 3002 stores the amount of money (reward fee) that the reputation information service operator should pay to the evaluation polarity reviewer and the amount of money (service fee) that should be received from the service user.
- the control unit 3001 has a function of calculating the amount of these fees and service usage charges and storing them in the money information storage means 3002.
- the reputation information service operator is a service provider that provides a distribution service of reputation information, and is an administrator of the reputation information service system 3000, the evaluation polarity estimation system 1000, and the reputation information extraction system 2000. .
- any one of the evaluation polarity estimation system 1000, the reputation information extraction system 2000, and the reputation information service system 3000, two or all of them can be handled as one information processing. It can be realized using a device! [0185] Next, the operation will be described. First, the operation of distributing reputation information to the service user terminal 5000 will be described.
- FIG. 19 is a flowchart showing an example of processing for distributing reputation information to the service user terminal 5000.
- the service user terminal 5000 inputs a natural text to extract reputation information according to the operation of the service user, and transmits it to the reputation information service system 3000 via the communication network (step S100). Then, the control unit 3001 of the reputation information service system 3000 receives natural text information from the service user terminal 5000 via the communication network.
- the control unit 3001 obtains reputation information from the natural text text. Specifically, the control unit 3001 transfers (sends) the natural text text received from the service user terminal 5000 to the reputation information extraction system 2000 via the communication network (step S101). Then, the reputation information extraction system 2000 extracts reputation information from the database based on the received natural text, and transmits it to the reputation information service system 3000 via the communication network (step S102).
- the control unit 3001 uses the evaluation polarity estimation system 1000 to input an evaluation expression, and obtains the evaluation polarity. Specifically, the control unit 3001 transfers (transmits) the reputation information received from the evaluation polarity estimation system 1000 to the evaluation polarity estimation system 1000 via the communication network (step S103). The evaluation polarity estimation system 1000 inputs (receives) reputation information and estimates the evaluation polarity (step S104) according to the same process as the evaluation polarity estimation process described in the first embodiment. The estimation result is returned to the reputation information service system 3 000.
- the evaluation polarity estimation system 1000 transmits the estimated evaluation polarity to the reputation information service system 3000 via the communication network (step S105), and the reputation information in the evaluation polarity estimation system 1000 is transmitted.
- Reputation information and its evaluation polarity are stored in the storage unit (reputation information storage unit 202).
- the evaluation polarity estimation system 1000 executes a process similar to the evaluation polarity estimation process described in the first embodiment. However, the evaluation polarity estimation system 1000 The evaluation polarity estimation process shown in the second embodiment or the third embodiment may be executed! / ⁇ ⁇ .
- the control unit 3001 transmits the reputation information extracted by the reputation information extraction system 2000 and the evaluation polarity for the reputation information estimated by the evaluation polarity estimation system 1000 to the service user terminal 5000 via the communication network. (Step S106). Then, the service user terminal 5000 presents the reputation information and its evaluation polarity to the service user. For example, the service user terminal 5000 displays the received reputation information and evaluation polarity on a display device such as a display device.
- control unit 3001 executes a billing process to the service user for using the reputation information distribution service (step S 107). Specifically, the control unit 3001 calculates the amount received by the service user power (service usage fee) and stores it in the money information storage means 3002. In this case, the control unit 3001 stores the money amount information and the identification information of the service user in the money information storage unit 3002 in association with each other.
- FIG. 20 is a flowchart showing an example of processing for reputing reputation information and evaluation polarity.
- the evaluation polarity predictor terminal 4000 inputs the object, attribute expression, and evaluation expression in order to search for reputation information to be browsed and reviewed according to the operation of the evaluation polarity investigator, and transmits it through the communication network. It is transmitted to the reputation information service system 3000 (step S200). Then, the control unit 3001 of the reputation information service system 3000 receives the object, the attribute expression, and the evaluation expression from the evaluation polarity predictor terminal 4000 via the communication network.
- the control unit 3001 Upon receiving the object, the attribute expression, and the evaluation expression, the control unit 3001 uses the reputation information reading means 111 of the evaluation polarity estimation system 1000 to receive the reputation information of the reputation information storage unit (reputation information storage unit 202). And its evaluation polarity are read out. Specifically, the control unit 3001 transmits the reputation information and the evaluation polarity extraction request together with the received object, attribute expression, and evaluation expression to the evaluation polarity estimation system 1000 via the communication network (step S2 01). ). Then, the reputation information reading means 111 of the evaluation polarity estimation system 1000 obtains the received object, attribute expression, reputation information that matches the evaluation expression, and evaluation polarity corresponding to the reputation information from the reputation information storage unit 202. Extract.
- the reputation information reading means 111 transmits the extracted reputation information and evaluation polarity to the reputation information service system 3000 via the communication network (step S202).
- the control unit 3001 transmits (transfers) the reputation information extracted by the evaluation polarity estimation system 1000 and the evaluation polarity to the guesser terminal 4000 via the communication network (step S203).
- the reviewer terminal 4000 receives the reputation information and its evaluation polarity via the communication network, and presents the evaluation information to the reviewer of the evaluation polarity to promote browsing and recommendation. For example, the evaluation polarity guesser terminal 4000 displays the received reputation information and evaluation polarity on a display device such as a display device.
- the evaluation polarity director browses the reputation information and the evaluation polarity, and operates the evaluation polarity director 400 to correct the incorrect reputation information and evaluation polarity.
- the evaluation polarity predictor terminal 4000 corrects the reputation information and the evaluation polarity in accordance with the operation of the evaluation polarity predictor, and transmits the corrected content to the reputation information service system 3000 via the communication network (step S204).
- the control unit 3001 of the reputation information service system 3000 transfers (transmits) the received corrected evaluation information and evaluation polarity to the evaluation polarity estimation system 1000 via the communication network (step S205). .
- the reputation information writing means 112 of the evaluation polarity estimation system 1000 stores the received corrected reputation information and evaluation polarity in the reputation information storage unit 202, and updates the stored contents of the reputation information storage unit 202 ( Step S206).
- control unit 3001 executes settlement processing for payment of the fee for evaluation to the polarity evaluation person for the evaluation of reputation information and evaluation polarity (step S207). Specifically, the control unit 3001 calculates information on the amount (reputation for consideration (reward fee)) to be paid by the reputation information service operator to the recommender and stores it in the money information storage means 3002. In this case, the control unit 3001 stores the money amount information and the identification information of the evaluation polarity guesser in the money information storage unit 3001 in association with each other.
- the service user and the evaluation polarity assessor may be the same. In that case, it is possible to eliminate the need to pay consideration to the evaluation polarity investigator (service user) or reduce the service usage fee that the service user should pay.
- the reputation information service system 3000 is extracted by the reputation information extraction system 2000 in response to a request from the service user terminal 5000.
- the evaluation polarity estimated by the evaluation polarity estimation system 1000 is distributed together with the reputation information.
- the accuracy of estimating the polarity of the other related reputation information is improved by adding one correct answer of the degree of polarity accumulated by the evaluation polarity estimation system (for every correct known polarity). Can be made. Therefore, as time elapses, it is possible to improve the estimation accuracy of the evaluation polarity with respect to reputation information while suppressing costs.
- the evaluation polarity is calculated based on the information accumulated in the reputation information accumulation unit. Therefore, it is possible to improve the estimation accuracy of the evaluation polarity of reputation information related to the reputation information as well as the reputation information suggested by the evaluation index predictor.
- the polarity estimation system is an evaluation polarity estimation system.
- the polarity estimation system is applied to applications that estimate polarity other than the evaluation polarity of reputation information. May be.
- the polarity estimation system may be applied to a purpose of estimating the polarity of a set of keywords (also referred to as a keyword set) extracted from various documents such as e-mail contents and information on an electronic bulletin board.
- Polarity is not limited to whether the information to be estimated is positive or negative, and when the keyword set to be estimated can be classified according to any two concepts! /, The power belonging to the concept of deviation, It may be information indicating! /.
- FIG. 21 is a block diagram showing a configuration example of the polarity estimation system in the fifth exemplary embodiment.
- the storage device 200 includes an expression storage unit 206 and an information storage unit 207 instead of the evaluation expression storage unit 201 and the reputation information storage unit 202. This is different from the embodiment.
- the basic functions of the constituent elements other than the expression storage unit 206 and the information storage unit 207 are the same as those functions described in the first embodiment. is there.
- the expression storage unit 206 stores in advance various expressions with known polarities.
- FIG. 22 is an explanatory diagram showing an example of various expressions and polarities stored in the expression storage unit 206.
- the expression storage unit 206 is a database that stores expressions and various polarities (polarities) in association with each other.
- the expression storage unit 206 stores a plurality of degrees of polarity in association with one expression.
- the polarities information indicating whether the expression indicates a full-fledged concept (also referred to as full-scale polarity) is used.
- full-scale polarity information indicating whether the expression indicates a full-fledged concept.
- the polarity of the full-scale extreme is close to “1”, and far from the full-fledged concept! /.
- the degree of heart warming polarity indicates that the heart warming atmosphere is expressed as the polarity degree is closer to “1”.
- the expression indicates an atmosphere in which the heart gets colder as the polarity of the heart warming polarity is closer to “1”.
- the degree of exhilaration polarity is closer to “1”, indicating that the expression indicates an atmosphere that makes you feel sick.
- the expression “natural nature” is a full-fledged concept and is an expression that indicates a heart-warming atmosphere. large.
- the expression “nature” is not an expression that expresses an atmosphere that makes you feel sick! /, So the value of the polarity of the refreshing polarity is small! /.
- the information storage unit 207 stores the keyword set and the degree of polarity output by the polarity estimation means 101.
- FIG. 23 shows an example of keyword set and polarity stored in the information storage unit 207. It is explanatory drawing shown.
- the information storage unit 207 is a database that stores a keyword set that can be included in various documents and various polarities of the keyword set in association with each other. In the present embodiment, as shown in FIG. 23, the information storage unit 207 stores a plurality of degrees of polarity in association with one keyword set as one record. Note that the keyword set and the degree of polarity stored in the information storage unit 207 are updated as needed based on the degree of polarity output by the polarity estimation means 101.
- the polarity estimation system estimates various polarities of the keyword set according to the same process as the evaluation polarity estimation system shown in the first embodiment estimates the reputation polarity of reputation information.
- the polarity estimation means 101 of the polarity estimation system inputs a keyword set to be estimated from the input means 300 according to the same processing as in step S10 shown in the first embodiment.
- the polarity estimation means 101 calculates various degrees of polarity for the keyword set to be estimated according to the same processing as steps S11 to S14 shown in the first embodiment.
- the polarity estimation means 101 causes the output means 400 to output the calculated various degrees of polarity in accordance with the same processing as Step S16 shown in the first embodiment.
- the polarity estimation means 101 For example, if there is a keyword in the keyword set that matches the expression stored in the expression storage unit 206 in accordance with the same processing as in step S11, the polarity estimation means 101, each corresponding to the expression. The degree of polarity is extracted from the expression storage unit 206.
- the polarity estimation unit 101 performs, for example, a keyword set among the records stored in the information storage unit 207 according to the same process as in step S13. The individual polarities are obtained by calculating the average value of the polarities of records that match the keywords in the middle. For example, in the example shown in FIG.
- the polarity estimation means 101 uses the keywords “golf”, “ground”, and the like among the records stored in the information storage unit 207, Extract all the records in which “Fighting spirit”, “Ball”, “Cloud”, “Arashi” and “Dream” appear, and calculate the average value of the polarities contained in those records
- the degree of polarity is calculated for each key word included in the information whose polarity is known.
- the key included in the information whose polarity is unknown Output the degree of polarity against the word. For this reason, it is possible to estimate various polarities by using information with known polarity for information with unknown polarity.
- the polarities of the keyword set may be estimated according to the same process as in the third embodiment or the third embodiment.
- the polarity estimation system may estimate various polarities of the keyword set by performing a predetermined weighting process in addition to the process shown in the present embodiment.
- the polarity estimation system may estimate various polarities of the keyword set based on the type of person who has determined the polarity of each keyword.
- the polarity estimation system may be applied to the use of a service model that distributes the polarity together with the keyword set in accordance with the same processing as in the fourth embodiment.
- an evaluation expression storage unit (for example, realized by the evaluation expression storage unit 201) that stores in advance an evaluation expression that is an expression indicating the evaluation of an object.
- the evaluation expression storage unit stores an evaluation expression polarity indicating whether the evaluation expression includes a positive expression or a negative expression, in association with the evaluation expression. Based on the evaluation expression stored in the evaluation expression storage unit and the evaluation expression polarity, the evaluation polarity of reputation information whose evaluation polarity is unknown may be estimated.
- the reputation information storage unit stores reputation information and evaluation polarity of reputation information in association with each other
- the polarity estimation means includes: Based on the reputation information and evaluation polarity stored in the reputation information storage unit, the evaluation polarity of reputation information whose evaluation polarity is unknown may be estimated.
- the reputation information storage unit obtains information at the time of acquisition (for example, the reputation information shown in FIG. 8). The acquired time) is stored in association with the reputation information, and the polarity estimation means determines the evaluation polarity corresponding to the reputation information stored in the reputation information storage unit based on the acquisition time information stored in the reputation information storage unit.
- the evaluation polarity is unknown based on the weighting means (for example, realized by the weighting means 1021), the evaluation polarity on which the weighting means performed weighting, and the reputation information stored in the reputation information storage unit. It can be used to estimate the evaluation polarity of certain reputation information.
- the reputation information storage unit evaluates evaluator information (for example, an evaluator ID) indicating an evaluator who evaluated the reputation information.
- the polarity estimation means estimates the evaluation polarity of reputation information whose evaluation polarity is unknown based on the reputation information and the evaluator information stored in the reputation information storage unit. Also good.
- the polarity estimation means includes a degree of polarity corresponding to an attribute expression included in reputation information whose evaluation polarity is known, and evaluation information.
- the degree of polarity corresponding to the included object and the degree of polarity corresponding to the evaluation expression included in the reputation information are calculated, and any one of the calculated degrees of polarity, or the power of each degree of polarity, Based on a combination of two or more, the total degree of polarity may be calculated by combining the degrees of polarity calculated for the input reputation information.
- the polarity estimation means corresponds to the polarity degree corresponding to the attribute expression, the polarity degree corresponding to the object, and the evaluation expression.
- the total degree of polarity may be obtained by calculating an average value, a total value, or a proportion of any one or two or more of the degrees of polarity.
- the polarity estimation means uses the attribute expression included in the input reputation information among the reputation information stored in the reputation information storage unit.
- the total degree of polarity corresponding to each piece of reputation information is calculated, the average of the degree of polarity corresponding to each piece of reputation information including the attribute expression included in the input reputation information is obtained, or the attribute expression included in the input reputation information is calculated.
- the tier calculates the sum of the polarities corresponding to each piece of reputation information including the object contained in the entered reputation information from among the reputation information stored in the reputation information storage unit, and selects the object contained in the entered reputation information. Even if the degree of polarity corresponding to an object is obtained by calculating the average of the degree of polarity corresponding to each included reputation information and calculating the ratio of reputation information including the object included in the input reputation information. Yo! /
- the polarity estimation means uses the evaluation expression included in the input reputation information among the reputation information stored in the reputation information storage unit.
- the total degree of polarity corresponding to each reputation information included is calculated, the average of the degree of polarity corresponding to each reputation information including the evaluation expression included in the entered reputation information is obtained, and the evaluation expression included in the entered reputation information is included.
- the polarity estimation unit may calculate the degree of polarity by weighting in order of time when reputation information is acquired.
- the polarity estimation unit calculates the degree of polarity according to the evaluator type indicating the type of the evaluator who evaluated the reputation information. It may be.
- the polarity estimation means includes the evaluator's age, gender, occupation, hobby, or purchased product as the evaluator type of reputation information. Depending on, the degree of polarity may be calculated.
- the polarity estimation unit is any one of the polarities corresponding to each keyword included in the information stored in the information storage unit.
- One, or any of them, may calculate the total polarity by calculating the average, sum or percentage of two or more! /.
- the polarity estimation means calculates the degree of polarity by weighting the information stored in the information storage unit in the order of acquisition time. Good.
- the stage may calculate the degree of polarity according to the evaluator type indicating the type of evaluator who evaluated the information stored in the information storage unit! /.
- the polarity estimation means includes the evaluator's age, sex, occupation, hobby as the evaluator type of information stored in the information storage unit. Or the degree of polarity may be calculated according to each purchased product! /.
- an evaluation expression storing step for storing in advance an evaluation expression that is an expression indicating evaluation of an object is included, and in the evaluation expression storing step, The evaluation expression polarity indicating whether the evaluation expression includes a positive expression or a negative expression is stored in association with the evaluation expression, and is stored based on the evaluation expression and the evaluation expression polarity stored in the polarity estimation step. Evaluation of reputation information whose evaluation polarity is unknown.
- reputation information and evaluation polarity of reputation information are stored in association with each other, and in the polarity estimation step Based on the stored reputation information and evaluation polarity! /, The evaluation polarity of reputation information whose evaluation polarity is unknown may be estimated.
- the reputation information storage step acquisition time information indicating when the reputation information is acquired is stored in association with the reputation information
- a predetermined weighting process is performed on the evaluation polarity corresponding to the stored reputation information based on the stored acquisition time information, and the evaluation polarity is determined based on the evaluation polarity subjected to the weighting process and the stored reputation information. It can be used to estimate the evaluation polarity of reputation information that is unknown.
- the evaluator information indicating the evaluator who evaluated the reputation information is stored in association with the reputation information
- the evaluation polarity of reputation information whose evaluation polarity is unknown may be estimated based on the stored reputation information and evaluator information.
- the degree of polarity corresponding to the attribute expression included in the reputation information whose evaluation polarity is known the evaluation information For evaluation expression included in the degree of polarity corresponding to the included object and reputation information
- the corresponding degree of polarity was calculated and calculated for the entered reputation information based on any one of the calculated degrees of polarity, or any power of each degree of polarity, a combination of two or more.
- the total degree of polarity may be calculated by combining the degree of polarity.
- the degree of polarity corresponding to the attribute expression, the degree of polarity corresponding to the object, and the evaluation expression are supported.
- the total degree of polarity may be obtained by calculating an average value, a total value, or a proportion of any one or two or more of the degrees of polarity.
- each evaluation information including the attribute expression included in the input reputation information is included.
- Calculate the sum of the corresponding polarities calculate the average of the polarities corresponding to each reputation information including attribute expressions included in the input reputation information, or the percentage of reputation information including attribute expressions included in the input reputation information It is also possible to calculate the degree of polarity corresponding to the attribute expression by calculating.
- the polarity estimation step in the polarity estimation step, it corresponds to each piece of reputation information including the object included in the input reputation information among the stored reputation information.
- the average degree of polarity corresponding to each reputation information including the object included in the input reputation information is calculated, and the ratio of the reputation information including the object included in the input reputation information is calculated.
- the degree of polarity corresponding to the object may be obtained.
- each evaluation information including the evaluation expression included in the input reputation information is included. Find the sum of the corresponding polarities, find the average of the polarities corresponding to each reputation information including the evaluation expressions included in the entered reputation information, and calculate the percentage of reputation information including the evaluation expressions included in the entered reputation information By calculating, the degree of polarity corresponding to the evaluation expression may be obtained.
- the polarity estimation may be performed by weighting in order of time when reputation information is acquired in the polarity estimation step.
- the degree of polarity is calculated according to the evaluator type indicating the type of the evaluator who evaluated the reputation information. It may be.
- the polarity estimation step in the polarity estimation step, as the evaluator type of reputation information, the evaluator's age, sex, occupation, hobby, or purchased product, respectively. Depending on, the degree of polarity may be calculated.
- the computer is caused to execute an evaluation expression storage process for preliminarily storing an evaluation expression that is an expression indicating evaluation of an object.
- the evaluation expression polarity indicating that the evaluation expression includes a positive expression or a negative expression is stored in association with the evaluation expression.
- a process for estimating the evaluation polarity of reputation information whose evaluation polarity is unknown may be executed.
- the reputation information and the evaluation polarity of the reputation information are stored in the computer in association with the reputation information storage process.
- the process may be executed to execute the process of estimating the evaluation polarity of reputation information whose evaluation polarity is unknown based on the stored reputation information and evaluation polarity in the polarity estimation process.
- the acquisition time information that indicates when the reputation information is acquired in the reputation information storage process in the computer is used as the reputation information. Based on the acquired acquisition time information in the polarity estimation process, a predetermined weighting process is executed on the evaluation polarity corresponding to the stored reputation information, and the weighting process is performed. Based on the evaluation polarity and the stored reputation information, it may be possible to execute a process for estimating the evaluation polarity of reputation information whose evaluation polarity is unknown.
- evaluator information indicating an evaluator who evaluated reputation information by reputation information storage processing in the computer is provided as evaluation information.
- evaluation information indicating an evaluator who evaluated reputation information by reputation information storage processing in the computer.
- the degree of polarity corresponding to the attribute expression included in the reputation information whose evaluation polarity is known in the polarity estimation process in the converter Execute the process of calculating the degree of polarity corresponding to the object included in the reputation information and the degree of polarity corresponding to the evaluation expression included in the reputation information. Based on the combination of two or more of the polarities, a process of calculating the total polarities by combining the polarities calculated for the input reputation information is executed.
- the polarity estimation process the polarity degree corresponding to the attribute expression, the polarity degree corresponding to the object. This is a process to calculate the total degree of polarity by calculating the average value, total value, or percentage of any one or two or more of the degrees of polarity corresponding to the expression. It ’s okay.
- the computer includes an attribute expression included in the input reputation information among the reputation information stored in the polarity estimation processing. Find the sum of the polarities corresponding to each piece of reputation information, calculate the average of the polarities corresponding to each piece of reputation information including the attribute expressions included in the entered reputation information, or include the attribute expressions contained in the entered reputation information By calculating the ratio of reputation information, a process for obtaining the degree of polarity corresponding to the attribute expression may be executed.
- the computer includes an object included in the input reputation information among the reputation information stored in the polarity estimation processing. Find the sum of the polarities corresponding to each reputation information, find the average of the polarities corresponding to each reputation information including the objects included in the entered reputation information, and include the objects included in the entered reputation information By calculating the ratio of reputation information, a process for obtaining the degree of polarity corresponding to the object may be executed.
- the polarity estimation program to which the present invention is applied it is included in the inputted reputation information among the reputation information stored in the computer by the polarity estimation processing. Obtain the total degree of polarity corresponding to each piece of reputation information including evaluation expression, obtain the average of the degree of polarity corresponding to each piece of reputation information including evaluation expression contained in the input reputation information, and evaluate the evaluation included in the inputted evaluation information You may perform the process which calculates
- the polarity is determined according to the evaluator type indicating the type of the evaluator who evaluated the reputation information in the polarity estimation process. You may perform the process which calculates a degree.
- the computer is used for the polarity estimation process, and the evaluator type of reputation information is the evaluator's age, gender, job, You may perform the process which calculates polarity according to each hobby or purchased goods.
- the present invention can be applied to the use of a service such as grasping an outline of good features and bad features of a product by determining the evaluation polarity of reputation information.
- the present invention is also applicable to a questionnaire automatic counting system.
- FIG. 1 is a block diagram showing an example of a configuration of a polarity estimation system according to an embodiment of the present invention.
- FIG. 2 is an explanatory diagram showing an example of evaluation expressions and evaluation polarities stored in an evaluation expression storage unit.
- FIG. 3 is an explanatory diagram showing examples of reputation information and evaluation polarities stored in a reputation information storage unit.
- FIG. 4 is an explanatory diagram showing another example of reputation information and evaluation polarity stored in the reputation information storage unit.
- FIG. 5 is a block diagram showing an example of the configuration of the polarity estimation means.
- FIG. 6 is a flowchart showing an example of processing for estimating the evaluation polarity by the evaluation polarity estimation system.
- FIG. 7 A block diagram showing an example of the configuration of the polarity estimation system according to the embodiment of the present invention.
- FIG. 9 is a block diagram showing an example of the configuration of polarity estimation means.
- FIG. 10 is a flowchart showing an example of processing for estimating the evaluation polarity by the evaluation polarity estimation system.
- FIG. 11 is a block diagram showing an example of the configuration of the polarity estimation system according to the embodiment of the present invention.
- FIG. 14 is a block diagram showing an example of the configuration of polarity estimation means.
- FIG. 15 is a flowchart showing an example of processing for estimating the evaluation polarity by the evaluation polarity estimation system.
- FIG. 16 is a block diagram showing a specific configuration example of an evaluation polarity estimation system.
- FIG. 17 is a block diagram showing an example of the configuration of the information service system according to the embodiment of the present invention.
- FIG. 18 A block diagram showing an example of the configuration of a polarity estimation system according to an embodiment of the present invention.
- FIG. 19 is a flowchart showing an example of a process for distributing reputation information to a service user terminal.
- FIG. 20 is a flowchart showing an example of processing for reputing reputation information and evaluation polarity.
- FIG. 21 is a block diagram showing an example of the configuration of the polarity estimation system according to the embodiment of the present invention.
- FIG. 22 It is explanatory drawing which shows an example of various expressions and polarity which an expression memory
- FIG. 23 It is explanatory drawing which shows an example of the keyword set and polarity which an information storage part memorize
Abstract
Description
Claims
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US12/448,010 US20100017391A1 (en) | 2006-12-18 | 2007-11-20 | Polarity estimation system, information delivery system, polarity estimation method, polarity estimation program and evaluation polarity estimatiom program |
JP2008550069A JP5151991B2 (ja) | 2006-12-18 | 2007-11-20 | 極性推定システム、情報配信システム、極性推定方法及び、極性推定用プログラム、及び評価極性推定用プログラム |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2006-340307 | 2006-12-18 | ||
JP2006340307 | 2006-12-18 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2008075524A1 true WO2008075524A1 (ja) | 2008-06-26 |
Family
ID=39536158
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2007/072484 WO2008075524A1 (ja) | 2006-12-18 | 2007-11-20 | 極性推定システム、情報配信システム、極性推定方法及び、極性推定用プログラム、及び評価極性推定用プログラム |
Country Status (4)
Country | Link |
---|---|
US (1) | US20100017391A1 (ja) |
JP (1) | JP5151991B2 (ja) |
CN (1) | CN101641693A (ja) |
WO (1) | WO2008075524A1 (ja) |
Cited By (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2010146171A (ja) * | 2008-12-17 | 2010-07-01 | Nippon Hoso Kyokai <Nhk> | 表現補完装置およびコンピュータプログラム |
JP2011070252A (ja) * | 2009-09-24 | 2011-04-07 | Hitachi Solutions Ltd | 文書解析システム |
JP2011159098A (ja) * | 2010-02-01 | 2011-08-18 | Nippon Telegr & Teleph Corp <Ntt> | 類似度計算装置、類似度計算方法および類似度計算プログラム |
CN102200969A (zh) * | 2010-03-25 | 2011-09-28 | 日电(中国)有限公司 | 基于句子顺序的文本情感极性分类系统和方法 |
JP2013175097A (ja) * | 2012-02-27 | 2013-09-05 | National Institute Of Information & Communication Technology | 述語テンプレート収集装置、特定フレーズペア収集装置、及びそれらのためのコンピュータプログラム |
WO2014065392A1 (ja) * | 2012-10-26 | 2014-05-01 | 日本電気株式会社 | 情報抽出システム、情報抽出方法および情報抽出用プログラム |
JP2015210700A (ja) * | 2014-04-28 | 2015-11-24 | Kddi株式会社 | 商品に対するユーザの感情分析装置及びプログラム |
US9268747B2 (en) | 2012-03-12 | 2016-02-23 | International Business Machines Corporation | Method for detecting negative opinions in social media, computer program product and computer |
US9740681B2 (en) | 2012-04-25 | 2017-08-22 | International Business Machines Corporation | Method for classifying pieces of text on basis of evaluation polarity, computer program product, and computer |
Families Citing this family (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9213687B2 (en) * | 2009-03-23 | 2015-12-15 | Lawrence Au | Compassion, variety and cohesion for methods of text analytics, writing, search, user interfaces |
US8600796B1 (en) * | 2012-01-30 | 2013-12-03 | Bazaarvoice, Inc. | System, method and computer program product for identifying products associated with polarized sentiments |
CN102821102B (zh) * | 2012-07-30 | 2016-09-21 | 中国电力科学研究院 | 一种智能配电网防御系统及其防御方法 |
US9633003B2 (en) * | 2012-12-18 | 2017-04-25 | International Business Machines Corporation | System support for evaluation consistency |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2006146567A (ja) * | 2004-11-19 | 2006-06-08 | Internatl Business Mach Corp <Ibm> | 表現検出システム、表現検出方法、及びプログラム |
Family Cites Families (21)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP1129417A4 (en) * | 1998-12-04 | 2004-06-30 | Technology Enabling Company Ll | DATA ORGANIZATION SYSTEMS AND METHODS |
US7275102B2 (en) * | 2001-01-22 | 2007-09-25 | Sun Microsystems, Inc. | Trust mechanisms for a peer-to-peer network computing platform |
GB2377782A (en) * | 2001-07-21 | 2003-01-22 | Ibm | Method and system for the communication of assured reputation information |
US7222187B2 (en) * | 2001-07-31 | 2007-05-22 | Sun Microsystems, Inc. | Distributed trust mechanism for decentralized networks |
JP2003058626A (ja) * | 2001-08-15 | 2003-02-28 | Fuji Photo Film Co Ltd | コンサルタントサーバ及びコンサルタント方法 |
AU2002342082A1 (en) * | 2001-10-18 | 2003-04-28 | Transpose, Llc | System and method for measuring rating reliability through rater prescience |
US7844610B2 (en) * | 2003-12-12 | 2010-11-30 | Google Inc. | Delegated authority evaluation system |
US7213047B2 (en) * | 2002-10-31 | 2007-05-01 | Sun Microsystems, Inc. | Peer trust evaluation using mobile agents in peer-to-peer networks |
US7769594B2 (en) * | 2003-09-05 | 2010-08-03 | France Telecom | Evaluation of reputation of an entity by a primary evaluation centre |
US20050216292A1 (en) * | 2003-10-11 | 2005-09-29 | Ashlock Jeffrey M | Method and system for financial evaluation of real estate properties |
US8200477B2 (en) * | 2003-10-22 | 2012-06-12 | International Business Machines Corporation | Method and system for extracting opinions from text documents |
JP3962382B2 (ja) * | 2004-02-20 | 2007-08-22 | インターナショナル・ビジネス・マシーンズ・コーポレーション | 表現抽出装置、表現抽出方法、プログラム及び記録媒体 |
JP2006004098A (ja) * | 2004-06-16 | 2006-01-05 | Internatl Business Mach Corp <Ibm> | 評価情報生成装置、評価情報生成方法、及びプログラム |
US20060230039A1 (en) * | 2005-01-25 | 2006-10-12 | Markmonitor, Inc. | Online identity tracking |
JP2007219880A (ja) * | 2006-02-17 | 2007-08-30 | Fujitsu Ltd | 評判情報処理プログラム、方法及び装置 |
US20080040748A1 (en) * | 2006-08-09 | 2008-02-14 | Ken Miyaki | Dynamic rating of content |
US8010511B2 (en) * | 2006-08-29 | 2011-08-30 | Attributor Corporation | Content monitoring and compliance enforcement |
US8782786B2 (en) * | 2007-03-30 | 2014-07-15 | Sophos Limited | Remedial action against malicious code at a client facility |
US8103727B2 (en) * | 2007-08-30 | 2012-01-24 | Fortinet, Inc. | Use of global intelligence to make local information classification decisions |
US20090265198A1 (en) * | 2008-04-22 | 2009-10-22 | Plaxo, Inc. | Reputation Evalution Using a contact Information Database |
US20110087613A1 (en) * | 2009-10-08 | 2011-04-14 | Evendor Check, Inc. | System and Method for Evaluating Supplier Quality |
-
2007
- 2007-11-20 US US12/448,010 patent/US20100017391A1/en not_active Abandoned
- 2007-11-20 WO PCT/JP2007/072484 patent/WO2008075524A1/ja active Search and Examination
- 2007-11-20 JP JP2008550069A patent/JP5151991B2/ja active Active
- 2007-11-20 CN CN200780051437A patent/CN101641693A/zh active Pending
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2006146567A (ja) * | 2004-11-19 | 2006-06-08 | Internatl Business Mach Corp <Ibm> | 表現検出システム、表現検出方法、及びプログラム |
Non-Patent Citations (3)
Title |
---|
HIROSHIMA N. ET AL.: "Hyoban Kensaku ni Okeru Query Izongata no Hyoka Kyokusei Fuyo (Query-sensitive Polarity Assignment for Sentiment Search)", INFORMATION PROCESSING SOCIETY OF JAPAN KENKYU HOKOKU, vol. 2006, no. 124, 23 November 2006 (2006-11-23), pages 129 - 134, XP003022913 * |
INUI K. ET AL.: "Text o Taisho to shita Hyoka Joho no Bunseki ni Kansuru Kenkyu Doko", JOURNAL OF NATURAL LANGUAGE PROCESSING, vol. 13, no. 3, 5 July 2006 (2006-07-05), pages 201 - 241, XP003022949 * |
KOBAYASHI N.: "Iken Chushutsu no Tame no Hyoka Hyogen no Shushu Collecting Evaluative Expressions for Opinion Extraction)", JOURNAL OF NATURAL LANGUAGE PROCESSING, vol. 12, no. 3, 10 July 2005 (2005-07-10), pages 203 - 222, XP003022922 * |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2010146171A (ja) * | 2008-12-17 | 2010-07-01 | Nippon Hoso Kyokai <Nhk> | 表現補完装置およびコンピュータプログラム |
JP2011070252A (ja) * | 2009-09-24 | 2011-04-07 | Hitachi Solutions Ltd | 文書解析システム |
JP2011159098A (ja) * | 2010-02-01 | 2011-08-18 | Nippon Telegr & Teleph Corp <Ntt> | 類似度計算装置、類似度計算方法および類似度計算プログラム |
CN102200969A (zh) * | 2010-03-25 | 2011-09-28 | 日电(中国)有限公司 | 基于句子顺序的文本情感极性分类系统和方法 |
JP2013175097A (ja) * | 2012-02-27 | 2013-09-05 | National Institute Of Information & Communication Technology | 述語テンプレート収集装置、特定フレーズペア収集装置、及びそれらのためのコンピュータプログラム |
WO2013128984A1 (ja) * | 2012-02-27 | 2013-09-06 | 独立行政法人情報通信研究機構 | 述語テンプレート収集装置、特定フレーズペア収集装置、及びそれらのためのコンピュータプログラム |
US9268747B2 (en) | 2012-03-12 | 2016-02-23 | International Business Machines Corporation | Method for detecting negative opinions in social media, computer program product and computer |
US9740681B2 (en) | 2012-04-25 | 2017-08-22 | International Business Machines Corporation | Method for classifying pieces of text on basis of evaluation polarity, computer program product, and computer |
WO2014065392A1 (ja) * | 2012-10-26 | 2014-05-01 | 日本電気株式会社 | 情報抽出システム、情報抽出方法および情報抽出用プログラム |
JPWO2014065392A1 (ja) * | 2012-10-26 | 2016-09-08 | 日本電気株式会社 | 情報抽出システム、情報抽出方法および情報抽出用プログラム |
JP2015210700A (ja) * | 2014-04-28 | 2015-11-24 | Kddi株式会社 | 商品に対するユーザの感情分析装置及びプログラム |
Also Published As
Publication number | Publication date |
---|---|
JPWO2008075524A1 (ja) | 2010-04-08 |
CN101641693A (zh) | 2010-02-03 |
US20100017391A1 (en) | 2010-01-21 |
JP5151991B2 (ja) | 2013-02-27 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP5151991B2 (ja) | 極性推定システム、情報配信システム、極性推定方法及び、極性推定用プログラム、及び評価極性推定用プログラム | |
US20230214941A1 (en) | Social Match Platform Apparatuses, Methods and Systems | |
US20200320604A1 (en) | Information Provision System, Information Provision Method, and Storage Medium | |
Vázquez et al. | A classification of user-generated content into consumer decision journey stages | |
US7546310B2 (en) | Expression detecting system, an expression detecting method and a program | |
Choi et al. | When seeing helps believing: The interactive effects of previews and reviews on e-book purchases | |
US8122371B1 (en) | Criteria-based structured ratings | |
AU2016264483A1 (en) | Method and system for effecting customer value based customer interaction management | |
JP6851894B2 (ja) | 対話システム、対話方法及び対話プログラム | |
US8452768B2 (en) | Using user search behavior to plan online advertising campaigns | |
US11461822B2 (en) | Methods and apparatus for automatically providing personalized item reviews | |
Rahmawati et al. | The effect of product quality, brand trust, price and sales promotion on purchase decisions on royal residence surabaya (Case Study In PT. Propnex Realti Visit) | |
Chen et al. | Studying product competition using representation learning | |
CN113407854A (zh) | 一种应用推荐方法、装置、设备及计算机可读存储介质 | |
CN110781428A (zh) | 评论展示方法、装置、计算机设备及存储介质 | |
Newberry et al. | Heterogeneous effects of online reputation for local and national retailers | |
US20210012405A1 (en) | Methods and apparatus for automatically providing personalized item reviews | |
Puspitasari et al. | Investigating the role of utilitarian and hedonic goals in characterizing customer loyalty in E-marketplaces | |
Nurmi et al. | Promotionrank: Ranking and recommending grocery product promotions using personal shopping lists | |
CN113868526A (zh) | 信息推送方法和装置、非瞬时性计算机可读存储介质 | |
US11487835B2 (en) | Information processing system, information processing method, and program | |
Anguera-Torrell et al. | The impact of positioning on click-through-rates in hotel metasearch engines | |
CN113971581A (zh) | 一种机器人的控制方法、装置、终端设备及存储介质 | |
CN113722487A (zh) | 用户情感分析方法、装置、设备及存储介质 | |
JP6809148B2 (ja) | プログラムおよび組み合わせ抽出システム |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
WWE | Wipo information: entry into national phase |
Ref document number: 200780051437.6 Country of ref document: CN |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 07832214 Country of ref document: EP Kind code of ref document: A1 |
|
DPE2 | Request for preliminary examination filed before expiration of 19th month from priority date (pct application filed from 20040101) | ||
WWE | Wipo information: entry into national phase |
Ref document number: 12448010 Country of ref document: US |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2008550069 Country of ref document: JP |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 07832214 Country of ref document: EP Kind code of ref document: A1 |
|
DPE2 | Request for preliminary examination filed before expiration of 19th month from priority date (pct application filed from 20040101) |