JP4501940B2 - Opinion extraction device, opinion extraction method, and opinion extraction program - Google Patents

Description

  The present invention relates to an opinion extraction device, an opinion extraction method, and an opinion extraction program for detecting an opinion expression part from an input text.

  For market analysis, the type of questionnaire survey selected from the options prepared for each question is easy to tabulate, and it is also possible to grasp trends and cross-analyze the relationship between questions. However, there is a limit to staying in the analysis from the viewpoint prepared in advance and digging deeper. To supplement this, it is effective to collect opinions in the form of free text. Opinions in free text format can be gathered by attaching to the questionnaire of the choice type. Recently, other than that, it is exchanged on an electronic bulletin board on the Internet, a personal homepage, a review site, or a mailing list. It has become possible to collect from various information sources such as e-mail.

  However, when gathering opinions from such diverse information sources, not only opinions about certain objects are written. That is, there is a problem that a sentence expressing an opinion and a sentence not being mixed are mixed. Therefore, it is necessary to have a technology that automatically detects only the points of opinion expression from the text. Furthermore, a technique is desired that can automatically perform the evaluation determination such as whether the extracted portion shows a positive opinion or a negative opinion.

  Therefore, as a conventional technique related to such an automatic extraction method of the opinion expression part, there is a technique for extracting an opinion in the form of a pair of a notation indicating an object and a notation indicating evaluation such as “good” and “bad” ( For example, Patent Document 1, Patent Document 2, Patent Document 3, Non-Patent Document 1, Non-Patent Document 2, Non-Patent Document 3, Non-Patent Document 4, and Non-Patent Document 5).

  For example, in the example of the opinion statement “I love mobile gear”, “mobile gear” is a notation indicating an object and “love” is a notation indicating evaluation.

  In the prior art, in the above example, the notation indicating the object is input to the system by the user, and the notation indicating the evaluation is stored in advance as a dictionary on the system side. Each notation indicating an evaluation registered in the dictionary is assigned an evaluation type of positive, negative, or neutral, and an opinion statement indicates whether the evaluation is positive, negative, or neutral. Is determined.

JP 2002-091981 A JP 2002-092004 A JP 2002-175330 A Kenji Tateishi, Yoshihide Ishiguro, Shunichi Fukushima, “Reputation Information Retrieval from the Internet”, Proceedings of the 62nd National Convention of Information Processing Society of Japan, 4W-5, 2001 Kenji Tateishi, Yoshihide Ishiguro, Shunichi Fukushima, "Prototype and Evaluation of Reputation Information Retrieval System", Proc. 63rd Annual Conference of Information Processing Society, 2V-1, 2001 Kenji Tateishi, Atsushi Morinaga, Kenji Yamanishi, Shunichi Fukushima, “Analysis of Opinions on the Web-Fusion of Information Extraction and Text Mining”, IPSJ 64th Annual Conference, 2X-4, 2002 Kenji Tateishi and Shunichi Fukushima, “Examination and Evaluation of Opinion Extraction Method in Opinion Analysis System”, Proceedings of the 1st Information Science and Technology Forum, D-1, 2002 Kenji Tateishi, Yoshihide Ishiguro, Shunichi Fukushima, “Reputation Information Retrieval from the Internet”, IPSJ Research Report, Natural Language Processing 144-11, 2001 Satoshi Morinaga, Kenji Yamanishi, Kenji Tateishi, Shunichi Fukushima, "Minning Product Reputations on the Web", The Eighth ACM SIGKDD International Discovery in D2

  The first problem of the prior art described above is that there is a limit to the determination accuracy of evaluation of positive, negative, or neutral.

  In the prior art, an evaluation type of affirmation, denial, or neutrality is assigned in advance to a notation indicating evaluation. FIG. 14 is an example of a dictionary used in a conventional opinion extraction device. In this dictionary, a notation 100 indicating evaluation is registered, and an evaluation type 200 of affirmation, denial, or neutrality is assigned to each notation 100 indicating evaluation.

  At this time, in the example of the notation indicating the evaluation registered in FIG. 14, “good” is positive and “bad” is negative. It is not always true that affirmation, “low” is negative. For example, when “high” is considered, it is considered positive if “performance is high” but negative if “price is high”. In other words, there are often cases where it is not possible to determine whether an evaluation is positive, negative, or neutral from the notation that simply represents the evaluation.

  In such a case, even in such a case, since the evaluation of whether it is affirmation, denial, or neutral is determined only from the notation that expresses the evaluation, an error is generated. The determination accuracy was limited.

  The second problem of the prior art is that even if the opinion evaluation (affirmation / denial / neutral) is determined, the basis of the evaluation cannot be accurately grasped. For example, when an affirmative opinion “I love mobile gear” is obtained, it is impossible to grasp what kind of nature of mobile gear is the basis for the affirmative opinion. It cannot be automatically extracted to the point that "low price" is the basis for the positive opinion, whether "high performance" or "the keyboard is easy to type" is the basis for the positive opinion. In order to conduct market analysis, it is very important to understand such a basis.

  For example, as in the technique described in Patent Document 2, there is a technique that classifies opinions based on evaluation points (for example, performance and price of an object). However, it was not possible to extract the basis of what kind of properties the object has at that point.

  In order to solve the conventional problems as described above, opinions are not simply seen as pairs of notation indicating the object and notation indicating the evaluation, but the structure of the opinion is modeled more precisely. It was necessary to handle.

  Therefore, the object of the present invention is based on a precise model of the structure of the opinion, and high accuracy can be obtained in the evaluation of the opinion evaluation of affirmation, denial or neutrality, and the nature of the object which is the basis for the opinion evaluation. Provide an opinion extraction device, an opinion extraction method, and an opinion extraction program.

  According to a first aspect of the invention for achieving the above object, a first notation indicating an object, a second notation indicating a focus of evaluation regarding the object, and an object of the object regarding the focus from an input text. Opinion characterized by comprising means for detecting a partial text including four elements of a third notation indicating properties and a fourth notation indicating positive or negative or neutral evaluation regarding the object as an opinion expressing part It is an extraction device.

  A second invention that achieves the above object is an opinion extraction device that detects an opinion expression part from an input text, and includes a first set of notations that can indicate an object, and a focus of evaluation on the object. A set of second notations that may be shown, a set of third notations that may indicate the nature of the object with respect to the point of interest, and a set of fourth notations that may indicate affirmative, negative, or neutral evaluation with respect to the object Based on the result of the opinion element storage means, the input text, the opinion element matching means for matching the set of notations registered in the opinion element storage means, and the opinion element matching means, An opinion expression location detecting means for detecting a partial text expressing an opinion from the input text is provided.

  A third invention that achieves the above object is an opinion extraction device that detects an opinion expression part from input text, and includes a first set of notations that can indicate an object, and a focus of evaluation on the object. A set of second notations that may be shown, a set of third notations that may indicate the nature of the object with respect to the point of interest, and a set of fourth notations that may indicate affirmative, negative, or neutral evaluation with respect to the object Based on the result of the opinion element storage means, the input text, the opinion element matching means for matching the set of notations registered in the opinion element storage means, and the opinion element matching means, Opinion expressed part detecting means for detecting a partial text expressing an opinion from the input text, and when the fourth notation is missing in the partial text detected by the opinion expressed part detecting means Rules for estimating positive, negative, or neutral evaluation from the property indicated by the third notation, or positive, negative, or neutral from the combination of the focus point indicated by the second notation and the property indicated by the third notation A positive, negative, or neutral evaluation is estimated from a rule that estimates the evaluation of the object, or a combination of the object indicated by the first notation, the focus point indicated by the second notation, and the property indicated by the third notation Evaluation estimation means for estimating an affirmative, negative, or neutral evaluation corresponding to the partial text detected by the opinion expression location detection means.

  A fourth invention for achieving the above object is an opinion extraction device for detecting an opinion expression part from an input text, wherein a first set of notations that can indicate an object and a focus of evaluation on the object are as follows. A set of second notations that may be shown, a set of third notations that may indicate the nature of the object with respect to the point of interest, and a set of fourth notations that may indicate affirmative, negative, or neutral evaluation with respect to the object Based on the result of the opinion element matching means, opinion element matching means for matching the input text with a set of notations registered in the opinion element storage means, and the input If the third notation is missing in the partial text detected by the opinion expressed part detecting means for detecting the partial text expressing the opinion from the text, and the opinion expressed part detecting means, A rule for estimating the property of an object from an affirmative, negative, or neutral evaluation indicated by the fourth notation, or an affirmation, negative, or neutral evaluation indicated by the second notation and the focus point indicated by the second notation A rule for estimating the property of an object from a combination of the above, or an object indicated by the first notation, a focus point indicated by the second notation, and a positive or negative or neutral evaluation indicated by the fourth notation. And a property estimation means for estimating the property of the object corresponding to the partial text detected by the opinion expression location detection means based on the rule for estimating the property of the object from the combination of the above.

  A fifth invention that achieves the above object is an opinion extraction device that detects an opinion expression part from an input text, and includes a first set of notations that can indicate an object, and a focus of evaluation on the object. A set of second notations that may be shown, a set of third notations that may indicate the nature of the object with respect to the point of interest, and a set of fourth notations that may indicate affirmative, negative, or neutral evaluation with respect to the object Based on the result of the opinion element storage means, the input text, the opinion element matching means for matching the set of notations registered in the opinion element storage means, and the opinion element matching means, Opinion expressed part detecting means for detecting a partial text expressing an opinion from the input text, and when the second notation is missing in the partial text detected by the opinion expressed part detecting means Rule for estimating the focus of evaluation from the property indicated by the third notation, or Rule for estimating the focus of evaluation from the combination of the object indicated by the first notation and the property indicated by the third notation Or based on a rule for estimating the focus of evaluation from the object indicated by the first notation and the positive, negative, or neutral evaluation indicated by the fourth notation. And a point-of-interest estimation means for estimating a point of focus of evaluation corresponding to the partial text.

  A sixth invention for achieving the above object is an opinion extraction device for detecting an opinion expression location from an input text, wherein (a) an object corresponding to each opinion expression location, (b) an evaluation relating to the object. And (c) an opinion element output means for outputting four items: (c) a property of the object relating to the eye point, and (d) a positive or negative or neutral evaluation relating to the object.

  According to a seventh aspect of the invention for achieving the above object, in the opinion extraction apparatus for detecting an opinion expression part from an input text, a first set of notations that can indicate an object and a focus of evaluation on the object can be indicated. A second set of notations, a third set of notations that may indicate the nature of the object with respect to the point of interest, and a fourth set of notations that may indicate a positive, negative, or neutral rating with respect to the object. Based on the result of the opinion element matching means, the opinion element matching means for checking the registered opinion element storage means, the input text and the set of notations registered in the opinion element storage means, and from the input text An opinion expression location detecting means for detecting a partial text expressing an opinion, and the fourth notation in the partial text detected by the opinion expression location detection means A rule for estimating a positive or negative or neutral evaluation from the property indicated by the third notation, or a positive or negative or neutral from the combination of the focus point indicated by the second notation and the property indicated by the third notation Affirmative, negative, or neutral evaluation is estimated from a rule that estimates evaluation, or a combination of an object indicated by the first notation, a focus point indicated by the second notation, and a property indicated by the third notation. Based on the rules, evaluation estimation means for estimating an affirmative, negative, or neutral evaluation corresponding to the partial text detected by the opinion expression location detection means; and the partial text detected by the opinion expression location detection means A rule for estimating the property of an object from the positive, negative, or neutral evaluation indicated by the fourth notation when the notation of 3 is missing, or A rule for estimating a property of an object from a combination of a focus point indicated by the second notation and an affirmation or negative or neutral evaluation indicated by the fourth notation, or the object indicated by the first notation and the The part detected by the opinion expression location detection means based on a rule for estimating the property of the object from the combination of the focus point indicated by the second notation and the positive or negative or neutral evaluation indicated by the fourth notation A property estimating means for estimating the property of the object corresponding to the text; and a property indicated by the third notation when the second notation is missing in the partial text detected by the opinion expression location detecting means. A rule for estimating an evaluation focus point from the above, or a rule for estimating an evaluation focus point from a combination of the object indicated by the first notation and the property indicated by the third notation, or Based on the rule that estimates the focus of evaluation from the object indicated by the first notation and the positive, negative, or neutral evaluation indicated by the fourth notation, the partial text detected by the opinion expression location detecting means Focusing point estimation means for estimating a corresponding point of evaluation, (a) an object corresponding to each opinion expression location, (b) an evaluation point regarding the object, (c) the object regarding the point of interest And (d) opinion element output means for outputting four items of positive, negative, or neutral evaluation regarding the object.

  According to an eighth aspect of the invention for achieving the above object, in the sixth or seventh aspect of the invention, the opinion element output means includes the items (a), (b), (c), and (d) in the partial text. It is characterized in that it is configured to output whether it is written or estimated.

  According to an eighth aspect of the invention for achieving the above object, in the second, third, fourth, fifth or seventh aspect of the invention, the opinion expression location detecting means includes: (1) the first The second notation, the second notation, the third notation, the partial text including all four elements of the fourth notation (2) among the four elements, the second notation, the third notation, Partial text including three elements lacking any of the fourth notation (3) Of the four elements, two elements lacking either the third notation or the fourth notation and the second notation According to the priority order of the partial text including the character string, the corresponding partial text is determined.

  A tenth invention that achieves the above object is characterized in that, in the second, third, fourth, fifth, or seventh invention, the partial text is a unit of a phrase, a sentence, or a paragraph.

  According to an eleventh aspect of the present invention that achieves the above object, the first notation indicating the object, the second notation indicating the focus of the evaluation related to the object, and the object related to the focus are selected from the input text. A partial text including four elements of a third notation indicating properties and a fourth notation indicating positive or negative or neutral evaluation regarding the object is detected as an opinion expression part.

  A twelfth aspect of the invention for achieving the above object is an opinion extraction method for detecting an opinion expression part from an input text, wherein a first set of notations that can indicate an object and a focus of evaluation on the object A set of second notations that may be shown, a set of third notations that may indicate the nature of the object with respect to the point of interest, and a set of fourth notations that may indicate affirmative, negative, or neutral evaluation with respect to the object Are previously registered in the opinion element memory, and the input element is collated with a set of notations registered in the opinion element memory, and the input based on the result of the opinion element matching step. An opinion expression detection step for detecting a partial text expressing an opinion from the text is executed.

  A thirteenth invention for achieving the above object is an opinion extraction method for detecting an opinion expression part from an input text, wherein a first set of notations that can indicate an object and a focus of evaluation on the object A set of second notations that may be shown, a set of third notations that may indicate the nature of the object with respect to the point of interest, and a set of fourth notations that may indicate affirmative, negative, or neutral evaluation with respect to the object Are previously registered in the opinion element memory, and the input element is collated with a set of notations registered in the opinion element memory, and the input based on the result of the opinion element matching step. An opinion expression location detection step for detecting a partial text expressing an opinion from the text, and the partial text detected by the opinion expression location detection step; Rule for estimating affirmation, negation, or neutral evaluation from the property indicated by the third notation, or the point of view indicated by the second notation and the property indicated by the third notation From the combination of the rule for estimating the positive or negative or neutral evaluation from the combination of the above, or the combination of the object indicated by the first notation, the attention point indicated by the second notation, and the property indicated by the third notation An evaluation estimating step for estimating an affirmative, negative, or neutral evaluation corresponding to the partial text detected by the opinion expression location detecting step based on a rule for estimating an affirmative, negative, or neutral evaluation; And

  A fourteenth aspect of the invention for achieving the above object is an opinion extraction method for detecting an opinion expression portion from an input text, wherein a first set of notations that can indicate an object and a focus of evaluation on the object are as follows. A set of second notations that may be shown, a set of third notations that may indicate the nature of the object with respect to the point of interest, and a set of fourth notations that may indicate affirmative, negative, or neutral evaluation with respect to the object Are previously registered in the opinion element memory, and the input element is collated with a set of notations registered in the opinion element memory, and the input based on the result of the opinion element matching step. An opinion expression location detecting step for detecting a partial text expressing an opinion from the text, and the third text in the partial text detected by the opinion expression location detection step A rule for estimating the property of an object from an affirmation or negative or neutral evaluation indicated by the fourth notation, or a focus point indicated by the second notation and the fourth notation A rule for estimating the property of an object from a combination of positive, negative, or neutral evaluations indicated by the above, or an object indicated by the first notation, an eye point indicated by the second notation, and the fourth notation A property estimation step for estimating the property of the object corresponding to the partial text detected by the opinion expression location detection step based on a rule for estimating the property of the object from a combination of an affirmative or negative or neutral evaluation shown; It is characterized by performing.

  According to a fifteenth aspect of the present invention for achieving the above object, there is provided an opinion extraction method for detecting an opinion expression part from an input text, wherein a first set of notations that can indicate an object, and a focus of evaluation on the object. A set of second notations that may be shown, a set of third notations that may indicate the nature of the object with respect to the point of interest, and a set of fourth notations that may indicate affirmative, negative, or neutral evaluation with respect to the object Are previously registered in the opinion element memory, and the input element is collated with a set of notations registered in the opinion element memory, and the input based on the result of the opinion element matching step. An opinion expression location detection step for detecting a partial text expressing an opinion from the text, and the second text in the partial text detected by the opinion expression location detection step A rule that estimates the focus of evaluation from the property indicated by the third notation or a combination of the object indicated by the first notation and the property indicated by the third notation Based on the rule for estimating the focus of evaluation from the above, or the rule for estimating the focus of evaluation from the positive or negative or neutral evaluation indicated by the object indicated by the first notation and the fourth notation, And a focus point estimation step of estimating a focus point of evaluation corresponding to the partial text detected by the opinion expression location detection step.

  According to a sixteenth aspect of the present invention for achieving the above object, there is provided an opinion extraction method for detecting an opinion expression portion from an input text, wherein (a) an object corresponding to each opinion expression portion, and (b) evaluation of the object. An opinion element output step for outputting four items, namely, a point of focus, (c) a property of the object related to the point of focus, and (d) an affirmation or negative or neutral evaluation regarding the object, is performed.

  According to a seventeenth aspect of the present invention for achieving the above object, there is provided an opinion extraction method for detecting an opinion expression portion from an input text, wherein a first set of notations that can indicate an object and a focus of evaluation on the object are as follows. A set of second notations that may be shown, a set of third notations that may indicate the nature of the object with respect to the point of interest, and a set of fourth notations that may indicate affirmative, negative, or neutral evaluation with respect to the object Are previously registered in the opinion element memory, and the input element is collated with a set of notations registered in the opinion element memory, and the input based on the result of the opinion element matching step. An opinion expression location detection step for detecting a partial text expressing an opinion from the text, and the partial text detected by the opinion expression location detection step; Rule for estimating affirmation, negation, or neutral evaluation from the property indicated by the third notation, or the point of view indicated by the second notation and the property indicated by the third notation From the combination of the rule for estimating the positive or negative or neutral evaluation from the combination of the above, or the combination of the object indicated by the first notation, the attention point indicated by the second notation, and the property indicated by the third notation Based on a rule for estimating a positive or negative or neutral evaluation, an evaluation estimation step for estimating a positive or negative or neutral evaluation corresponding to the partial text detected by the opinion expression detection step, and the opinion expression detection Positive or negative or neutral evaluation indicated by the fourth notation when the third notation is missing in the partial text detected by the step A rule for estimating the property of the object, or a rule for estimating the property of the object from the combination of the focus point indicated by the second notation and the positive or negative or neutral evaluation indicated by the fourth notation, or Based on the rule for estimating the property of the object from the combination of the object indicated by the first notation, the focus point indicated by the second notation, and the positive, negative, or neutral evaluation indicated by the fourth notation The second notation is missing in the property estimation step for estimating the property of the object corresponding to the partial text detected by the opinion expression location detection step and the partial text detected by the opinion expression location detection step. A rule for estimating the point of evaluation from the property indicated by the third notation, or the object indicated by the first notation and the property indicated by the third notation Based on the rule for estimating the evaluation point from the combination, or the rule for estimating the evaluation point from the object indicated by the first notation and the positive or negative or neutral evaluation indicated by the fourth notation, A focus point estimation step for estimating a focus point of evaluation corresponding to the partial text detected by the opinion expression location detection step, (a) an object corresponding to each opinion expression location, and (b) evaluation of the object. And (c) an opinion element output step of outputting four items: (d) a property of the object related to the point of interest, and (d) an affirmation or negative or neutral evaluation regarding the object. .

  According to an eighteenth aspect of the invention for achieving the above object, in the sixteenth or seventeenth aspect, the opinion element output step includes the steps (a), (b), (c), and (d) in the partial text. It is characterized in that it is output by distinguishing whether it is written or estimated.

  According to a nineteenth aspect of the invention for achieving the above object, in the twelfth, thirteenth, fourteenth, fifteenth or seventeenth aspect of the invention, the opinion expression location detecting step includes: (1) the first The second notation, the second notation, the third notation, the partial text including all four elements of the fourth notation (2) among the four elements, the second notation, the third notation, Partial text including three elements lacking any of the fourth notation (3) Of the four elements, two elements lacking either the third notation or the fourth notation and the second notation In this case, the corresponding partial text is determined in accordance with the priority order of the partial text including.

  A twentieth invention that achieves the above object is characterized in that, in the twelfth, thirteenth, fourteenth, fifteenth or seventeenth invention, the partial text is a unit of a phrase, a sentence or a paragraph.

  According to a twentieth aspect of the invention for achieving the above object, there is provided an opinion extraction program for causing an information processing apparatus to detect an opinion expression part from an input text, the first notation indicating an object from the input text; , A second notation indicating the focus of evaluation on the object, a third notation indicating the nature of the object regarding the focus, and a fourth indicating an affirmative, negative, or neutral evaluation regarding the object This is an opinion extraction program for executing a process of detecting a partial text including four elements of notation as an opinion expression part.

  According to a twenty-second aspect of the invention for achieving the above object, there is provided an opinion extraction program for causing an information processing apparatus to detect an opinion expression portion from input text, wherein the opinion extraction program is registered in the information processing apparatus in a memory. A first set of notations that can indicate the object, a second set of notations that can indicate the focus of evaluation on the object, and a third notation that can indicate the properties of the object regarding the focus Based on the result of the opinion element matching process that matches the set and a fourth notation set that may indicate a positive or negative or neutral evaluation of the object, and the input text, An opinion expression location detection process for detecting a partial text expressing an opinion from the input text is executed.

  According to a twenty-third aspect of the invention for achieving the above object, there is provided an opinion extraction program for causing an information processing apparatus to detect an opinion expression part from input text, wherein the opinion extraction program is registered in the information processing apparatus in a memory. A first set of notations that can indicate the object, a second set of notations that can indicate the focus of evaluation on the object, and a third notation that can indicate the properties of the object regarding the focus Based on the result of the opinion element matching process that matches the set and a fourth notation set that may indicate a positive or negative or neutral evaluation of the object, and the input text, An opinion expression location detection process for detecting a partial text expressing an opinion from the input text, and the fourth text in the partial text detected by the opinion expression location detection process. Rules for estimating affirmative, negative, or neutral evaluation from the properties indicated by the third notation when the notation is missing, or the points indicated by the second notation and the properties indicated by the third notation From the combination of the rule for estimating the positive or negative or neutral evaluation from the combination of the above, or the combination of the object indicated by the first notation, the attention point indicated by the second notation, and the property indicated by the third notation Based on a rule for estimating a positive, negative, or neutral evaluation, an evaluation estimation process for estimating a positive, negative, or neutral evaluation corresponding to the partial text detected by the opinion expression location detection process is executed. And

  According to a twenty-fourth aspect of the invention for achieving the above object, there is provided an opinion extraction program for causing an information processing apparatus to detect an opinion expression portion from input text, wherein the opinion extraction program is registered in the information processing apparatus in a memory. A first set of notations that can indicate the object, a second set of notations that can indicate the focus of evaluation on the object, and a third notation that can indicate the properties of the object regarding the focus Based on the result of the opinion element matching process that matches the set and a fourth notation set that may indicate a positive or negative or neutral evaluation of the object, and the input text, An opinion expression location detection process for detecting a partial text expressing an opinion from the input text, and the third text in the partial text detected by the opinion expression location detection process. A rule for estimating a property of an object from an affirmation or negative or neutral evaluation indicated by the fourth notation, or a focus point indicated by the second notation and the fourth notation A rule for estimating the property of an object from a combination of positive, negative, or neutral evaluations indicated by the above, or an object indicated by the first notation, an eye point indicated by the second notation, and the fourth notation A property estimation process for estimating the property of the object corresponding to the partial text detected by the opinion expression location detection process, based on a rule for estimating the property of the object from a combination of a positive or negative or neutral evaluation to be indicated; Is executed.

  According to a twenty-fifth aspect of the present invention for achieving the above object, there is provided an opinion extraction program for causing an information processing apparatus to detect an opinion expression part from input text, wherein the opinion extraction program is registered in the information processing apparatus in a memory. A first set of notations that can indicate the object, a second set of notations that can indicate the focus of evaluation on the object, and a third notation that can indicate the properties of the object regarding the focus Based on the result of the opinion element matching process that matches the set and a fourth notation set that may indicate a positive or negative or neutral evaluation of the object, and the input text, An opinion expression location detection process for detecting a partial text expressing an opinion from the input text, and the partial text detected by the opinion expression location detection process If the notation is missing, the rule for estimating the focus of evaluation from the property indicated by the third notation, or the object indicated by the first notation and the property indicated by the third notation Based on the rule for estimating the evaluation point from the combination, or the rule for estimating the evaluation point from the object indicated by the first notation and the positive or negative or neutral evaluation indicated by the fourth notation, And a focus point estimation process for estimating a focus point of evaluation corresponding to the partial text detected by the opinion expression location detection process.

  According to a twenty-sixth aspect of the present invention for achieving the above object, there is provided an opinion extraction program for causing an information processing apparatus to detect an opinion expression part from input text, wherein the opinion extraction program causes the information processing apparatus to express each opinion expression. (A) the object corresponding to the location, (b) the focus of the evaluation regarding the object, (c) the nature of the object regarding the focus, (d) the positive or negative or neutral evaluation regarding the object, This is an opinion extraction program for executing opinion element output processing for outputting the four items.

  According to a twenty-seventh aspect of the invention for achieving the above object, there is provided an opinion extraction program for causing an information processing apparatus to detect an opinion expression part from input text, wherein the opinion extraction program is registered in the information processing apparatus in a memory. A first set of notations that can indicate the object, a second set of notations that can indicate the focus of evaluation on the object, and a third notation that can indicate the properties of the object regarding the focus Based on the result of the opinion element matching process that matches the set and a fourth notation set that may indicate a positive or negative or neutral evaluation of the object, and the input text, An opinion expression location detection process for detecting a partial text expressing an opinion from the input text, and the fourth text in the partial text detected by the opinion expression location detection process. Rules for estimating affirmative, negative, or neutral evaluation from the properties indicated by the third notation when the notation is missing, or the points indicated by the second notation and the properties indicated by the third notation From the combination of the rule for estimating the positive or negative or neutral evaluation from the combination of the above, or the combination of the object indicated by the first notation, the attention point indicated by the second notation, and the property indicated by the third notation Based on a rule for estimating an affirmative, negative, or neutral evaluation, an evaluation estimation process for estimating an affirmative, negative, or neutral evaluation corresponding to the partial text detected by the opinion expression detection process, and the opinion expression detection When the third notation is missing in the partial text detected by the processing, the object from the positive or negative or neutral evaluation indicated by the fourth notation A rule for estimating a property, or a rule for estimating a property of an object from a combination of a focus point indicated by the second notation and an affirmation or negative or neutral evaluation indicated by the fourth notation, or the first The opinion expression based on a rule for estimating the property of the object from the combination of the object indicated by the notation, the focus point indicated by the second notation, and the positive, negative, or neutral evaluation indicated by the fourth notation When the second notation is missing in the partial text detected by the property estimation processing for estimating the property of the object corresponding to the partial text detected by the location detection processing and the opinion expression location detection processing, A rule for estimating the point of interest from the property indicated by the third notation, or a combination of the object indicated by the first notation and the property indicated by the third notation. Based on a rule for estimating an eye point, or a rule for estimating an eye point of evaluation from an object indicated by the first notation and an affirmative, negative or neutral evaluation indicated by the fourth notation A focus point estimation process for estimating a focus point of evaluation corresponding to the partial text detected by the detection process, (a) a target object corresponding to each opinion expression location, (b) a focus point of evaluation regarding the target object, ( and (c) an opinion element output process for outputting four items, i.e., a property of the object related to the point of interest, and (d) an affirmation or negative or neutral evaluation regarding the object.

  In a twenty-eighth aspect of the present invention that achieves the above object, in the above-described twenty-sixth or twenty-seventh aspect, the opinion element output processing includes: It is characterized in that it is output by distinguishing whether it is written or estimated.

  In a twenty-ninth aspect of the present invention that achieves the above object, in the twenty-second, twenty-third, twenty-fourth, twenty-fifth, or twenty-seventh aspect, the opinion expression location detection process includes: (1) the first The second notation, the second notation, the third notation, the partial text including all four elements of the fourth notation (2) among the four elements, the second notation, the third notation, Partial text including three elements lacking any of the fourth notation (3) Of the four elements, two elements lacking either the third notation or the fourth notation and the second notation In this case, the corresponding partial text is determined in accordance with the priority order of the partial text including.

  A thirtieth invention for achieving the above object is characterized in that, in the twenty-second, twenty-third, twenty-fourth, twenty-fifth or twenty-seventh invention, the partial text is a unit of a phrase, sentence or paragraph.

  A thirty-first invention for achieving the above object is an opinion extraction device for detecting an opinion expression part from an input text, and includes a first set of notations that can indicate an object and a focus of evaluation on the object. A set of second notations that may be shown, a set of third notations that may indicate the nature of the object with respect to the point of interest, and a set of fourth notations that may indicate affirmative, negative, or neutral evaluation with respect to the object Based on the result of the opinion element storage means, the input text, the opinion element matching means for matching the set of notations registered in the opinion element storage means, and the opinion element matching means, In the input text, an opinion statement is the first notation, the second notation, the third notation, and the fourth notation, and the opinion statement is the first notation, Second notation and The second composition pattern consisting of the third notation, the third composition pattern consisting of the first notation, the second notation and the fourth notation, and the opinion expression being the first notation A fourth composition pattern comprising the notation, the third notation, and the fourth notation, a fifth composition pattern comprising the opinion expression comprising the first notation and the third notation, and an opinion expression comprising the first composition pattern. And an opinion expression location detecting means for detecting a partial text corresponding to any one of the first notation and the sixth constituent pattern consisting of the fourth notation.

  In a thirty-second invention that achieves the above object, in the thirty-first invention, when the opinion expression detection unit detects a partial text from the input text, the first pattern is set as a first priority. The second configuration pattern, the third configuration pattern, and the fourth configuration pattern have a second priority, and the fifth configuration pattern and the sixth configuration pattern have a third priority. It is configured to detect from the highest ranking.

  In a thirty-third invention that achieves the above object, in the above thirty-first or thirty-second invention, the partial text detected by the opinion expression location detecting means is the second configuration pattern or the fifth configuration pattern. A rule for estimating an affirmative, negative, or neutral evaluation from the property indicated by the third notation in the partial text, or the point of view indicated by the second notation and the property indicated by the third notation in the partial text Rules for estimating positive, negative, or neutral evaluation from the combination of the above, or the object indicated by the first notation, the focus point indicated by the second notation, and the nature indicated by the third notation in the partial text Based on the rule that estimates affirmation, denial, or neutral evaluation from the combination of It is characterized by further having an evaluation estimating means for estimating the evaluation of negative or neutral.

  In a thirty-fourth aspect of the present invention that achieves the above object, in any one of the thirty-first to thirty-third aspects, the partial text detected by the opinion expression location detecting means is a third configuration pattern or a sixth configuration pattern. The rule for estimating the property of the object from the positive, negative, or neutral evaluation indicated by the fourth notation in the partial text, or the focus point indicated by the second notation and the first point in the partial text. A rule for estimating the property of an object from a combination of positive, negative, or neutral evaluation indicated by 4 or the object indicated by the first notation and a focus point indicated by the second notation in a partial text And the opinion expression location detection means based on a rule for estimating the property of the object from the combination of positive or negative or neutral evaluation indicated by the fourth notation. Characterized in that it further has the property estimating means for estimating the properties of the object corresponding to the partial text.

  In a thirty-fifth aspect of the present invention that achieves the above object, in any one of the thirty-first to thirty-fourth aspects, the partial text detected by the opinion expression location detecting means is a fourth configuration pattern, a fifth configuration pattern, or In the case of the sixth configuration pattern, the rule for estimating the focus of evaluation from the property indicated by the third notation in the partial text, or the object indicated by the first notation in the partial text and the first From the rule for estimating the point of evaluation from the combination with the property indicated by 3 or from the object indicated by the first notation in the partial text and the positive or negative or neutral evaluation indicated by the fourth notation Based on the rule for estimating the evaluation point of interest, the method further comprises a point of interest estimation unit for estimating the point of evaluation of the evaluation corresponding to the partial text detected by the opinion expression location detection unit. The features.

  In a thirty-sixth aspect of the present invention that achieves the above object, in any one of the thirty-first to thirty-fifth aspects, (a) an object corresponding to each opinion expression location, (b) an evaluation point regarding the object, ( It further comprises opinion element output means for outputting four items: c) the nature of the object related to the point of interest, and (d) positive or negative or neutral evaluation related to the object.

  In a thirty-seventh aspect of the invention that achieves the above object, in the thirty-sixth aspect of the invention, the opinion element output means includes the items (a), (b), (c), and (d) described in the partial text. It is characterized by distinguishing whether it is a thing or the estimated thing and outputting it.

  A thirty-eighth aspect of the present invention that achieves the above object is an opinion extraction program that causes an information processing device to detect an opinion expression location from input text, wherein the opinion extraction program specifies the information processing device as a target. A set of first notation that can be shown, a set of second notation that can indicate the focus of evaluation on the object, a set of third notation that can indicate the nature of the object with respect to the focus, Opinion element storage means registered with a fourth set of notations that can indicate positive or negative or neutral evaluation of the object, and the input text and the set of notations registered in the opinion element storage means are collated Based on the result of the opinion element matching means and the opinion element matching means, an opinion statement is expressed from the input text in the first notation, the second notation, the third notation, and the first notation. A first composition pattern consisting of the following notations, an opinion statement is the first notation, a second composition pattern consisting of the second notation and the third notation, and an opinion statement is the first notation, A third composition pattern comprising a second notation and the fourth notation; a fourth composition pattern comprising an opinion statement comprising the first notation, the third notation and the fourth notation; and an opinion expression. Is a portion corresponding to any one of the fifth composition pattern consisting of the first notation and the third notation and the sixth composition pattern whose opinion expression is composed of the first notation and the fourth notation. It is made to function as an opinion expression location detection means for detecting text.

  In a thirty-ninth aspect of the present invention that achieves the above object, in the thirty-eighth aspect, when the opinion expression location detecting means detects a partial text from the input text, the first pattern is set as a first priority. The second configuration pattern, the third configuration pattern, and the fourth configuration pattern have a second priority, and the fifth configuration pattern and the sixth configuration pattern have a third priority. It is an opinion extraction program characterized by making it function so that it may detect from a thing with high order.

  In a fortieth aspect of the invention for achieving the above object, in the thirty-eighth or thirty-ninth aspect, the opinion extraction program is the information processing device, and the partial text detected by the opinion expression location detecting means is the second configuration. In the case of the pattern or the fifth constituent pattern, a rule for estimating a positive or negative or neutral evaluation from the property of the third notation in the partial text, or the indication of the second notation in the partial text Rules for estimating affirmative, negative, or neutral evaluation from the combination of the point of view and the property indicated by the third notation, or the object indicated by the first notation and the second notation indicated in the partial text Based on a rule for estimating affirmation, denial, or neutral evaluation from the combination of the point of view and the property indicated by the third notation, Characterized in that to further function as an evaluation estimating means for estimating a positive or evaluation of negative or neutral corresponding to the detected partial text I.

  In a forty-first aspect of the present invention that achieves the above object, in any one of the thirty-eighth to forty-first aspects, the opinion extraction program uses the information processing apparatus, and the partial text detected by the opinion expression location detecting means 3 component pattern or the sixth component pattern, the rule for estimating the property of the object from the positive or negative or neutral evaluation indicated by the fourth notation in the partial text, or the above-mentioned in the partial text A rule for estimating the property of an object from a combination of a focus point indicated by the second notation and an affirmation or negative or neutral evaluation indicated by the fourth notation, or an object indicated by the first notation in a partial text Based on a rule that estimates the property of an object from a combination of an object, a focus point indicated by the second notation, and an affirmative, negative, or neutral evaluation indicated by the fourth notation Te, characterized in that to further function as a property estimating means for estimating the properties of the object corresponding to the detected partial text by the opinion portion detecting means.

  According to a forty-second invention for achieving the above object, in any one of the thirty-eighth to forty-first inventions, the opinion extraction program uses the information processing apparatus, and the partial text detected by the opinion expression location detecting means When the configuration pattern is 4, the fifth configuration pattern, or the sixth configuration pattern, the rule for estimating the focus of evaluation from the property of the third notation in the partial text, or the above-mentioned in the partial text A rule for estimating a focus of evaluation based on a combination of the object indicated by the first notation and the property indicated by the third notation, or the object indicated by the first notation in the partial text and the fourth Based on the rule for estimating the focus of the evaluation from the positive, negative, or neutral evaluation indicated by the notation, the evaluation corresponding to the partial text detected by the opinion expression location detecting means Characterized in that to further function as Viewpoints estimating means for estimating the eye point.

  In a forty-third invention for achieving the above object, in any one of the thirty-eighth to forty-second inventions, the opinion extraction program causes the information processing apparatus to correspond to each opinion expression location (a) object, (b Further, as an opinion element output means for outputting four items: (e) a point of evaluation regarding the object, (c) a property of the object regarding the point of interest, and (d) a positive, negative, or neutral evaluation regarding the object. It is made to function.

  In a forty-fourth aspect of the present invention that achieves the above object, in the above-mentioned forty-third aspect, the opinion element output means includes the items (a), (b), (c), and (d) described in the partial text. It is an opinion extraction program characterized in that it functions so as to distinguish and output whether it is a thing or an estimated one.

  As described above, in the present invention, an opinion is expressed as (a) an object, (b) a focus of evaluation regarding the object, (c) a property of the object regarding the focus, (d) a positive or negative or neutral evaluation regarding the object. , And are configured to detect notation corresponding to these four elements. In addition, if there is something missing, it is estimated from other factors.

  This has an excellent effect that high accuracy can be obtained in the determination of the evaluation of an opinion of affirmation, denial or neutrality. In addition, it is possible to extract up to the nature of the object that is the basis for opinion evaluation.

  A first embodiment of the present invention will be described with reference to the drawings.

  FIG. 1 is a block diagram showing the configuration of the first embodiment.

  In FIG. 1, 1 is a text input means. The text input unit 1 is a unit for inputting a text to be subjected to opinion extraction processing. For example, a keyboard, an OCR, a pen input, a voice recognition device, a unit for reading a text file placed on a network, etc. Anything that can input text is acceptable.

  A text buffer 2 records the text input by the text input means 1.

  Reference numeral 3 denotes opinion element storage means. This opinion element storage means 3 can indicate the first set of notations that can indicate the object, the second set of notations that can indicate the focus of evaluation on the object, and the properties of the object regarding the focus. A third set of notations and a fourth set of notations that may indicate a positive or negative or neutral rating for the object are stored.

  FIG. 2 is a diagram illustrating an example of contents stored in the opinion element storage unit 3. As shown in FIG. 2, the opinion element storage unit 3 includes a first notation storage unit 30, a second notation storage unit 31, a third notation storage unit 32, and a fourth notation storage unit 33. It has four parts. The conventional opinion extraction apparatus includes only the first notation storage unit 30 and the fourth notation storage unit 33, whereas the opinion element storage unit 3 of the present invention has the second notation storage. The unit 31 and the storage unit 32 of the third notation are provided and are characterized by being composed of four parts.

  The second notation storage unit 31, the third notation storage unit 32, and the fourth notation storage unit 33 are created and registered in advance in a dictionary sense. On the other hand, the storage unit 30 having the first notation is configured to set the object selected by the user each time.

  Further, the second notation storage unit 31 and the third notation storage unit 32 store notation appropriate for each field such as information processing and food, and the object (first item) selected by the user. By selecting the field to which (notation of 1) belongs, the accuracy in discriminating opinion elements can be improved. However, it is not essential that the second notation storage unit 31 and the third notation storage unit 32 be configured separately for each field as described above.

  In FIG. 2, as an example of each recording unit, in the information processing field, liquid crystal, keyboard, performance, price, design, etc. are registered in the second notation storage unit 31, and in the third notation storage unit 32, Pretty, high, low ... are registered. In the food field, taste, fragrance, safety, etc. are registered in the second notation storage unit 31, and clean, delicious, tasteless, etc. are registered in the third notation storage unit 32.

  4 is an opinion element collating means. This opinion element collating means 4 collates the input text with a set of notations registered in the opinion element storage means 3. Then, the collation result buffer 5 records where the above four types of notations appear in the input text output from the text buffer 2.

  6 is an opinion expression part detection means. The opinion expression location detection unit 6 detects a partial text expressing an opinion from the input text based on the result of the opinion element verification unit 4 output from the verification result buffer 5. Here, the partial text means, for example, a sentence unit, and includes variations such as a phrase or paragraph. The partial text detected by the opinion expression location detection means 6 is output to the detection result buffer 7 and recorded in the detection result buffer 7.

  Although it is a detection method of the opinion expression location detection means 6, in the present invention, as described above, in the present invention, the opinion is expressed by (a) the object, (b) the point of evaluation regarding the object, and (c) the object regarding the point of interest. (D) positive or negative or neutral evaluation of the object. In view of this, in the opinion element storage unit 3, the configuration patterns that can be taken by these four elements are registered.

  FIG. 3 describes an opinion composition pattern based on these four types of elements. There are six opinion composition patterns [A] [B] [C] [D] [E] [F] shown in FIG.

  The composition pattern [A] is, for example, “(a object) + (b) the evaluation point regarding the object b” + (c) Property) + (positive or negative or neutral evaluation of the d object). In FIG. 3, it is abbreviated as (a) + (b) + (c) + (d).

  The composition pattern [B] is, for example, “(a object) + (the point of interest for evaluation regarding the object b) + (the property of the object regarding the point of eye c)” as “the taste of this spaghetti is delicious”. It is a configuration. In FIG. 3, it is abbreviated as (a) + (b) + (c).

  The composition pattern [C] is, for example, “I like the taste of this spaghetti.” (A object) + (point of evaluation regarding b object) + (positive or negative or neutral regarding object d) (Evaluation). In FIG. 3, it is abbreviated as (a) + (b) + (d).

  The composition pattern [D] is, for example, (a object) + (the property of the object with respect to the c point of interest) + (a positive or negative or neutral with respect to the object), such as “I love this spaghetti because it is delicious.” (Evaluation). In FIG. 3, it is abbreviated as (a) + (c) + (d).

  The configuration pattern [E] is, for example, a configuration of (a target object) + (subject property regarding the point of c focus), such as “This spaghetti is delicious.” In FIG. 3, it is abbreviated as (a) + (c).

  The configuration pattern [F] is, for example, a configuration of (a object) + (positive or negative or neutral evaluation regarding the d object) as “I love this spaghetti”. In FIG. 3, it is abbreviated as (a) + (d).

  The opinion expression location detection means 6 extracts an opinion expression location according to the six configuration patterns.

  Next, the operation of extracting an opinion expression part will be described.

  FIG. 4 is an example of a flowchart of the opinion expression location detection means 6. In the flowchart shown in FIG. 4, an opinion expression part is extracted in the following procedure.

  First, in step 101, a partial text (sentence) corresponding to the opinion composition pattern [A] is detected. That is, a partial text including all four elements of the first notation (a), the second notation (b), the third notation (c), and the fourth notation (d) is detected.

  Next, in step 102, a partial text (sentence) corresponding to the opinion composition pattern [B] [C] [D] is detected from the text in the range not detected in step 601. That is, among the four elements, a partial text including three elements lacking any of the second notation (b), the third notation (c), and the fourth notation (d) is detected.

  Finally, in step 103, a partial text (sentence) corresponding to the opinion composition pattern [E] [F] is detected from the text in the range not detected in steps 601 and 602. That is, of the four elements, a partial text including two elements lacking either the third notation (c) or the fourth notation (d) and the second notation (b) is detected.

  Next, the operation of an example corresponding to the first embodiment will be described using a specific example of input text.

  FIG. 5 shows an example of input text written in the text buffer 2. The content of the opinion element storage means 3 is the same as that illustrated in FIG. 2, the object written in the first notation storage unit 30 is “spaghetti”, the second notation storage unit 31 and the second notation It is assumed that the food field is selected for the storage unit 32 with the number 3.

  At this time, when the opinion element extraction means 4 is executed, a result as shown in FIG. In FIG. 6, the first notation (a), the second notation (b), the third notation (c), and the fourth notation (d) indicate which part of the input text matches. Yes. In FIG. 6, the matching text in the input text is shown in a format in which a matching part is marked, but the data format of the matching result buffer 5 is expressed in a format in which position information corresponding to a list of matched notations is added. Multiple variations are possible.

  Using this result, the opinion expression location detection means 6 detects an opinion expression location according to the flowchart shown in FIG.

  The result is as shown in FIG. From the input text as an example, 4 opinions were found. 7, # 1 corresponds to the opinion composition pattern [E], # 2 corresponds to the opinion composition pattern [A], # 3 corresponds to the opinion composition pattern [C], and # 4 corresponds to the opinion composition pattern [D].

  Thus, by extracting the opinion by distinguishing the four elements that constitute the opinion, it is possible to capture the content of the opinion more accurately. In the conventional opinion extraction method, the element (c) that is the property of the object and the element (d) that indicates the evaluation of the object are handled without being distinguished, but it is possible to distinguish and grasp these. . Further, since the point of interest (b) and the property (c) are also extracted at the same time, it is possible to grasp the basis of the evaluation.

  Next, a second embodiment of the present invention will be described with reference to the drawings.

  FIG. 8 is a block diagram showing the configuration of the second embodiment.

  The seven constituent elements of the text input means 1, the text buffer 2, the opinion element storage means 3, the opinion element collation means 4, the collation result buffer 5, the opinion expression location detection means 6, and the detection result buffer 7 are the first embodiment. And in common.

  In the second embodiment, four constituent elements are newly added: a viewpoint estimation unit 8, a property estimation unit 9, an evaluation estimation unit 10, and an opinion element output unit 11. Hereinafter, these four components will be described.

  When the second notation is missing in the partial text detected by the opinion expression location detecting means 6, the point-of-interest estimation unit 8 evaluates the point of evaluation (b) from the property (c) indicated by the third notation. From the combination of the rule [(c) → (b)] for estimating the object (a) indicated by the first notation and the property (c) indicated by the third notation, the evaluation point (b) From the rule to be estimated [(a) + (c) → (b)] or the object (a) indicated by the first notation and the positive or negative or neutral evaluation (d) indicated by the fourth notation Based on the rule [(a) + (d) → (b)] for estimating the point of interest (b), the point of interest of evaluation corresponding to the partial text detected by the opinion expression location detection unit 6 is estimated.

  FIG. 9 shows an example of the eye point estimation rule used by the eye point estimation means 8. The estimated one is enclosed in a square.

  As described above, the following three evaluation estimation rules can be considered. (C) → (b) type (a) + (c) → (b) type (a) + (d) → (b) type For example, the rules “c” → (b) type “delicious” → “Taste” is estimated to be “taste” based on the property of “delicious”. When the property is “thin”, it can be estimated that the point of interest is “thickness” in many cases, but not only that, but the case of the point of interest being “taste” is also conceivable. Therefore, using the rule of (a) + (c) → (b), depending on what the object is, “book” + “thin” → “thickness” or “beer” + “thin” "→" Taste ". On the other hand, depending on the object, there are cases where the attention point is narrowed down in many cases. With the (a) + (d) → (b) type rule, the attention point is estimated from the object and the evaluation. In the case of matching a plurality of rules, it is necessary to determine the priority order between the rules. For example, (a) + (c) → (b) type, (c) → (b) type , (A) + (d) → (b) types may be preferentially applied.

  When the third notation is missing in the partial text detected by the opinion expression location detecting means 6, the property estimating means 9 determines the object from the positive or negative or neutral evaluation (d) indicated by the fourth notation. Rule [(d) → (c)] for estimating the property (c) of the above, or the focus point (b) indicated by the second notation and the positive or negative or neutral evaluation (d) indicated by the fourth notation [(B) + (d) → (c)] for estimating the property (c) of the object from the combination of the object (a) indicated by the first notation and the point of focus indicated by the second notation Rule [(a) + (b) + (d) → (c) for estimating the property (c) of the object from the combination of (b) and the positive or negative or neutral evaluation (d) indicated by the fourth notation )]), The pair corresponding to the partial text detected by the opinion expression location detection means 6 To estimate the nature of things.

  FIG. 10 shows an example of the property estimation rule used by the property estimation means 9. The estimated one is enclosed in a square.

  As described above, the following three evaluation estimation rules can be considered. (D) → (c) type (b) + (d) → (c) type (a) + (b) + (d) → (c) type For example, (d) → (c) type rule. When “affirmation” → “interesting” is an opinion of “affirmation”, the property of the object is estimated to be “interesting”. In addition to being “affirmed”, if it is known that the focus of evaluation is “price”, “price” + “affirmation” → “cheap” (b) + (d) → (C) A type rule is used.

  In the case of “affirmation”, when the focus is “performance”, the rule “performance” + “affirmation” → “high” can be applied. Furthermore, in addition to the evaluation and focus, if it is known what the object is, the rule of (a) + (b) + (d) → (c) type is applied. In this way, a rule that matches a larger number of elements is preferentially applied.

  When the fourth notation is missing in the partial text detected by the opinion expression location detecting means 6, the evaluation estimating means 10 evaluates positive or negative or neutral (from the property (c) indicated by the third notation ( d) is estimated from the combination of the rule [(c) → (d)] or the point of view (b) indicated by the second notation and the property (c) indicated by the third notation. Rule [(b) + (c) → (d)] for estimating the evaluation (d), or the object (a) indicated by the first notation and the focused point (b) indicated by the second notation and the third Based on the rule [(a) + (b) + (c) → (d)] for estimating a positive or negative or neutral evaluation from the combination with the property (c) indicated by Positive or negative or neutral rating corresponding to the partial text detected by Estimated to.

  FIG. 11 shows an example of the evaluation estimation rule used by the evaluation estimation means 10. The estimated one is enclosed in a square.

  As described above, the following three evaluation estimation rules can be considered. (C) → (d) type (b) + (c) → (d) type (a) + (b) + (c) → (d) type For example, (c) → (d) type rule. “Clean” → “Affirmation” estimates the evaluation of “Affirmation” from the property of the object “Clean”. If the property of the object is “High”, it cannot be determined whether it is “Positive” or “Negative”, and “Price” + “High” → “Negative”, “Performance” + “High” → As in the case of “Yes”, it is necessary to make a determination with a set point of interest, and a rule of (b) + (c) → (d) type is used. In addition, the (a) + (b) + (c) → (d) type rule further determines the evaluation taking into consideration that the evaluation is not determined only by the focus and nature, but what the target is. This is the case. In the case of matching a plurality of rules, a rule with a larger number of elements may be applied with priority.

  The estimation results by the point of interest estimation means 8, the property estimation means 9, and the evaluation estimation means 10 are additionally recorded in the detection result buffer 7.

  The opinion element output means 11 includes (a) an object corresponding to each opinion expression location, (b) a focus of evaluation regarding the object, (c) a property of the object regarding the focus, (d) affirmation regarding the object, or Outputs four items, negative or neutral evaluation. Depending on where each opinion is expressed, all four of these elements may be listed, or some may be missing. In the case where the element is missing, those that can be estimated by the evaluation estimating means 10, the property estimating means 9, and the focused point estimating means 8 output the estimation results. It should be noted that elements that are originally described in the sentence or estimated ones may be displayed and output separately. For example, for the estimated element, a mark indicating the meaning of estimation may be given. Alternatively, there is a method of writing a score indicating the reliability of estimation together.

  Next, the operation of an example corresponding to the second embodiment will be described using a specific example of input text.

  This example is the same as the example related to the first embodiment until the point where the opinion is expressed is detected. Therefore, the continuation from where the result of FIG. 7 is obtained in the detection result buffer 7 will be described.

  In the result of FIG. 7, four opinion statements are detected. At this point, the satisfaction status of the four opinion elements in each opinion statement is as follows.

  # 1: Element (b) and element (d) are missing.

  # 2: All four elements are included, and there are no missing elements.

  # 3: Element (c) is missing.

  # 4: Element (b) is missing.

  The missing elements are estimated by the evaluation estimation unit 10, the property estimation unit 9, and the eye point estimation unit 8. In this estimation process, missing elements are estimated for each opinion statement, and this is repeated in order for each opinion statement, and all opinion statements are estimated for each type of missing element. Although the method of repeating for the element type can be considered, here, the former procedure will be described.

  Regarding the opinion statement # 1 “spaghetti eaten at that time” in FIG. 7, the elements of the point of interest (b) and the evaluation (d) are not described. Therefore, the eye point estimation means 8 uses the (c) → (b) type eye point estimation rule based on the property (c) of “delicious” and the eye point (b) is “taste”. Estimated. Further, the evaluation estimation means 10 estimates that the evaluation (d) is “affirmed” using the (c) → (d) type evaluation estimation rule based on the property (c) “delicious”. .

  Since the opinion statement # 2 has four elements, the element (c) is not written in the opinion statement # 3 “I hate the taste of that store's spaghetti.” Therefore, the property estimation means 9 uses the (b) + (d) → (c) type based on the point of interest (b) “taste” and the evaluation (d) “denial” (= “dislike”). Using the property estimation rule, it is estimated that the property (c) is “tasteful”.

  Regarding the opinion statement # 4, “I love that spaghetti because it is delicious,” the element of the point of interest (b) is not written. Therefore, as in the case of # 1, the viewpoint estimation means 8 uses the viewpoint estimation rule of (c) → (b) based on the property (c) of “delicious” to determine the viewpoint ( It is estimated that b) is “taste”.

  As a result of such estimation processing, the contents of the detection result buffer 7 are as shown in FIG. Finally, the opinion element output means 11 outputs the contents of FIG.

  As described above, in this embodiment, when any of the four elements constituting the opinion is missing, it is possible to estimate it from the other elements. This makes it possible to accurately estimate the basis for the opinion. In addition, since a mechanism for determining a positive, negative, or neutral evaluation from a combination of a plurality of elements is used, the determination accuracy can be improved as compared with the conventional method.

  In contrast to the first embodiment, in the second embodiment, four constituent elements, namely, a spot estimation means 8, a property estimation means 9, an evaluation estimation means 10, and an opinion element output means 11 are added. In addition to the form in which all these four components are added, there may be a form in which only a part of the four components is added. In that case, although some of the opinion elements cannot be estimated, there is still an effect that the content of the opinion can be grasped more accurately than before.

  A third embodiment of the present invention will be described with reference to the drawings.

  FIG. 13 is a block diagram illustrating a configuration of the third embodiment.

  The third embodiment includes an input device 20, a data processing device 21, a storage device 22, and an output device 25, and further for realizing the opinion extraction device of the first and second embodiments described above. A recording medium 26 on which the program is recorded is provided. The recording medium 26 may be a magnetic disk, semiconductor memory, CD-ROM, or other recording medium.

  The input device 20 is a device for inputting instructions from an operator, such as a mouse and a keyboard. The output device 25 is a device that outputs a processing result by the data processing device 21 such as a display screen or a printer.

  A program for realizing the opinion extraction device is read from the recording medium 26 into the data processing device 21, controls the operation of the data processing device 21, and generates an input memory 23 and a work memory 24 in the storage device 22.

  The data processing device 21 controls the text input unit 1, the opinion element collating unit 4, the opinion expression location detecting unit 6, and the second in the first and second embodiments by controlling a program for realizing the opinion extracting device. The same processing as that of the point-of-interest estimation means 8, the property estimation means 9, the evaluation estimation means 10 and the like in the embodiment is executed.

FIG. 1 is a block diagram showing the configuration of the first embodiment of the opinion extraction apparatus of the present invention. FIG. 2 is a diagram showing an example of the contents of the opinion element storage means. FIG. 3 is a diagram showing an opinion composition pattern and an example thereof. FIG. 4 is a flowchart of the opinion expression location detection means. FIG. 5 is a diagram showing an example of the contents of the text buffer. FIG. 6 is a diagram showing an example of the contents of the collation result buffer. FIG. 7 shows an example of the contents of the detection result buffer. FIG. 8 is a block diagram showing the configuration of the second embodiment of the opinion extraction apparatus of the present invention. FIG. 9 is a diagram illustrating an example of the attention point estimation rule. FIG. 10 is a diagram illustrating an example of the property estimation rule. FIG. 11 is a diagram illustrating an example of the evaluation estimation rule. FIG. 12 is a diagram illustrating an example of the contents of the detection result buffer after estimation of missing elements. FIG. 13 is a block diagram showing the configuration of the third embodiment of the opinion extraction apparatus of the present invention. FIG. 14 is a diagram showing an example of a dictionary of expressions indicating evaluation used in a conventional opinion extracting apparatus.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Text input means 2 Text buffer 3 Opinion element memory | storage means 4 Opinion element collation means 5 Collation result buffer 6 Opinion expression part detection means 7 Detection result buffer 20 Input device 21 Data processing device 22 Storage device 23 Input memory 24 Work memory 25 Output device 26 Recording Medium 30 First Notation Storage Unit 31 Second Notation Storage Unit 32 Third Notation Storage Unit 33 Fourth Notation Storage Unit

Claims (23)

An opinion extraction program for causing an information processing device to detect an opinion expression part from input text,
The opinion extraction program is stored in the information processing apparatus.
A first set of notations registered in the memory that can indicate the object, a second set of notations that can indicate the focus of evaluation on the object, and a third that can indicate the properties of the object regarding the focus And a character string corresponding to the first to fourth notations from the input text based on a set of notations and a fourth notation set that may indicate a positive, negative, or neutral evaluation of the object. Matching process,
Based on the result of the opinion element matching process, a partial text that expresses an opinion from the input text, and an opinion statement location detection process that detects the first to fourth notations in a distinguished state;
A rule for storing the property indicated by the third notation and the positive / negative / neutral evaluation in association with each other when the fourth notation is missing in the partial text detected by the opinion expression location detection process Or a rule for storing a combination of the focus point indicated by the second notation and the property indicated by the third notation and an evaluation of affirmation, negation, or neutrality , or the indication of the first notation The opinion expression location detection process based on a rule that stores a combination of a target object, a focus point indicated by the second notation, and a property indicated by the third notation, and an affirmative, negative, or neutral evaluation. An opinion extraction program for executing an evaluation estimation process for estimating an affirmative, negative, or neutral evaluation corresponding to the partial text detected by. An opinion extraction program for causing an information processing device to detect an opinion expression part from input text,
The opinion extraction program is stored in the information processing apparatus.
A first set of notations registered in the memory that can indicate the object, a second set of notations that can indicate the focus of evaluation on the object, and a third that can indicate the properties of the object regarding the focus And a string of characters corresponding to the first to fourth notations from the input text based on a set of notations and a fourth notation set that may indicate a positive, negative, or neutral rating on the object Opinion element matching process,
Based on the result of the opinion element matching process, a partial text that expresses an opinion from the input text, and an opinion statement location detection process that detects the first to fourth notations in a distinguished state;
When the third notation is missing in the partial text detected by the opinion expression location detection processing, the positive or negative or neutral evaluation indicated by the fourth notation is associated with the property of the object. storing rules, or the second rule in association with the nature of the combination and the object of the evaluation of the positive or negative, or neutral showing the Viewpoints of said fourth representation indicated by the notation, or the A rule that stores a combination of the object indicated by the first notation, the focus point indicated by the second notation, the positive, negative, or neutral evaluation indicated by the fourth notation and the property of the object in association with each other . An opinion extraction program for executing a property estimation process for estimating a property of an object corresponding to the partial text detected by the opinion expression location detection process based on the above. An opinion extraction program for causing an information processing device to detect an opinion expression part from input text,
The opinion extraction program is stored in the information processing apparatus.
A first set of notations registered in the memory that can indicate the object, a second set of notations that can indicate the focus of evaluation on the object, and a third that can indicate the properties of the object regarding the focus And a string of characters corresponding to the first to fourth notations from the input text based on a set of notations and a fourth notation set that may indicate a positive, negative, or neutral rating on the object Opinion element matching process,
Based on the result of the opinion element matching process, a partial text that expresses an opinion from the input text, and an opinion statement location detection process that detects the first to fourth notations in a distinguished state;
A rule for storing the property indicated by the third notation and the focus of evaluation in association with each other when the second notation is missing in the partial text detected by the opinion expression location detection process, or A rule for storing a combination of an object indicated by the first notation and a property indicated by the third notation and a focus point of evaluation , or an object indicated by the first notation and the fourth The focus of evaluation corresponding to the partial text detected by the opinion expression location detection process based on the rule that stores the combination of the positive or negative or neutral evaluation indicated by the notation and the focus of evaluation in association with each other Opinion extraction program for executing point of interest estimation processing for estimating An opinion extraction program for causing an information processing device to detect an opinion expression part from input text,
The opinion extraction program is stored in the information processing apparatus.
A first set of notations registered in the memory that can indicate the object, a second set of notations that can indicate the focus of evaluation on the object, and a third that can indicate the properties of the object regarding the focus And a string of characters corresponding to the first to fourth notations from the input text based on a set of notations and a fourth notation set that may indicate a positive, negative, or neutral rating on the object Opinion element matching process,
Based on the result of the opinion element matching process, a partial text that expresses an opinion from the input text, and an opinion statement location detection process that detects the first to fourth notations in a distinguished state;
A rule for storing the property indicated by the third notation and the positive / negative / neutral evaluation in association with each other when the fourth notation is missing in the partial text detected by the opinion expression location detection process Or a rule for storing a combination of the focus point indicated by the second notation and the property indicated by the third notation and an evaluation of affirmation, negation, or neutrality , or the indication of the first notation The opinion expression location detection process based on a rule that stores a combination of a target object, a focus point indicated by the second notation, and a property indicated by the third notation, and an affirmative, negative, or neutral evaluation. A rating estimation process that estimates a positive, negative, or neutral rating corresponding to the partial text detected by
When the third notation is missing in the partial text detected by the opinion expression location detection processing, the positive or negative or neutral evaluation indicated by the fourth notation is associated with the property of the object. storing rules, or the second rule in association with the nature of the combination and the object of the evaluation of the positive or negative, or neutral showing the Viewpoints of said fourth representation indicated by the notation, or the A rule that stores a combination of the object indicated by the first notation, the focus point indicated by the second notation, the positive, negative, or neutral evaluation indicated by the fourth notation and the property of the object in association with each other . Based on the property estimation processing for estimating the property of the object corresponding to the partial text detected by the opinion expression location detection processing,
A rule for storing the property indicated by the third notation and the focus of evaluation in association with each other when the second notation is missing in the partial text detected by the opinion expression location detection process, or A rule for storing a combination of an object indicated by the first notation and a property indicated by the third notation and a focus point of evaluation , or an object indicated by the first notation and the fourth The focus of evaluation corresponding to the partial text detected by the opinion expression location detection process based on the rule that stores the combination of the positive or negative or neutral evaluation indicated by the notation and the focus of evaluation in association with each other Focusing point estimation processing for estimating
(A) the object corresponding to each opinion expression location, (b) the focus of evaluation regarding the object, (c) the nature of the object regarding the focus, (d) affirmation or denial or neutrality regarding the object An opinion extraction program for executing an opinion element output process for outputting four items of evaluation. In the opinion element output process, the items (a), (b), (c), and (d) are output by distinguishing whether each item is described in the partial text or estimated. The opinion extraction program according to claim 4, wherein the opinion extraction program is performed. The opinion expression location detection processing includes: (1) partial text including all four elements of the first notation, the second notation, the third notation, and the fourth notation from the input text ( 2) Of the four elements, partial text including three elements lacking any of the second notation, the third notation, and the fourth notation. (3) Of the four elements, the third notation The corresponding partial text is determined according to the priority order of the partial text including two elements lacking either the notation or the fourth notation and the second notation. 6. The opinion extraction program according to any one of items 5. The opinion extraction program according to any one of claims 1 to 6, wherein the partial text is a unit of a phrase, a sentence, or a paragraph. An opinion extraction device that detects an opinion expression from input text,
A set of first notation that can indicate the object, a set of second notation that can indicate the focus of evaluation on the object, and a set of third notation that can indicate the properties of the object regarding the focus And an opinion element storage means in which is registered a fourth notation set that can indicate a positive, negative, or neutral evaluation of the object;
Opinion element matching means for matching a character string corresponding to the first to fourth notations registered in the opinion element storage means from the input text;
Based on the result of the opinion element matching means, an opinion statement location detecting means for detecting a partial text expressing an opinion from the input text in a state where the first to fourth notations are distinguished,
A rule for storing the property indicated by the third notation and the positive / negative / neutral evaluation in association with each other when the fourth notation is missing in the partial text detected by the opinion expression location detecting means Or a rule for storing a combination of the focus point indicated by the second notation and the property indicated by the third notation and an evaluation of affirmation, negation, or neutrality , or the indication of the first notation Based on a rule that stores a combination of an object, a focus point indicated by the second notation and a property indicated by the third notation, and an affirmative, negative, or neutral evaluation , the opinion expression location detecting means And an evaluation estimation means for estimating an affirmative, negative, or neutral evaluation corresponding to the partial text detected by. An opinion extraction device that detects an opinion expression from input text,
A set of first notation that can indicate the object, a set of second notation that can indicate the focus of evaluation on the object, and a set of third notation that can indicate the properties of the object regarding the focus And an opinion element storage means in which is registered a fourth notation set that can indicate a positive, negative, or neutral evaluation of the object;
Opinion element matching means for matching a character string corresponding to the first to fourth notations registered in the opinion element storage means from the input text;
Based on the result of the opinion element matching means, an opinion statement location detecting means for detecting a partial text expressing an opinion from the input text in a state where the first to fourth notations are distinguished,
When the third notation is missing in the partial text detected by the opinion expression location detecting means, the positive or negative or neutral evaluation indicated by the fourth notation is associated with the property of the object. storing rules, or the second rule in association with the nature of the combination and the object of the evaluation of the positive or negative, or neutral showing the Viewpoints of said fourth representation indicated by the notation, or the A rule that stores a combination of the object indicated by the first notation, the focus point indicated by the second notation, the positive, negative, or neutral evaluation indicated by the fourth notation and the property of the object in association with each other . An opinion extraction apparatus comprising: property estimation means for estimating the property of the object corresponding to the partial text detected by the opinion expression location detection means. An opinion extraction device that detects an opinion expression from input text,
A set of first notation that can indicate the object, a set of second notation that can indicate the focus of evaluation on the object, and a set of third notation that can indicate the properties of the object regarding the focus And an opinion element storage means in which is registered a fourth notation set that can indicate a positive, negative, or neutral evaluation of the object;
Opinion element matching means for matching a character string corresponding to the first to fourth notations registered in the opinion element storage means from the input text;
Based on the result of the opinion element matching means, an opinion statement location detecting means for detecting a partial text expressing an opinion from the input text in a state where the first to fourth notations are distinguished,
A rule for storing the property indicated by the third notation and the focus of evaluation in association with each other when the second notation is missing in the partial text detected by the opinion expression location detecting unit, or A rule for storing a combination of an object indicated by the first notation and a property indicated by the third notation and a focus point of evaluation , or an object indicated by the first notation and the fourth Based on a rule that stores a combination of positive, negative, or neutral evaluations indicated by the notation and an evaluation focus in association with each other, the evaluation focus corresponding to the partial text detected by the opinion expression location detection means An opinion extracting apparatus comprising: a point of interest estimating means for estimating In an opinion extraction device that detects an opinion expression from input text,
A set of first notation that can indicate the object, a set of second notation that can indicate the focus of evaluation on the object, and a set of third notation that can indicate the properties of the object regarding the focus And an opinion element storage means that registers a fourth set of notations that may indicate a positive, negative, or neutral evaluation of the object;
Opinion element matching means for matching a character string corresponding to the first to fourth notations registered in the opinion element storage means from the input text;
Based on the result of the opinion element matching means, an opinion statement location detecting means for detecting a partial text expressing an opinion from the input text in a state where the first to fourth notations are distinguished,
A rule for storing the property indicated by the third notation and the positive / negative / neutral evaluation in association with each other when the fourth notation is missing in the partial text detected by the opinion expression location detecting means Or a rule for storing a combination of the focus point indicated by the second notation and the property indicated by the third notation and an evaluation of affirmation, negation, or neutrality , or the indication of the first notation Based on a rule that stores a combination of an object, a focus point indicated by the second notation and a property indicated by the third notation, and an affirmative, negative, or neutral evaluation , the opinion expression location detecting means A rating estimator for estimating a positive or negative or neutral rating corresponding to the partial text detected by
When the third notation is missing in the partial text detected by the opinion expression location detecting means, the positive or negative or neutral evaluation indicated by the fourth notation is associated with the property of the object. storing rules, or the second rule in association with the nature of the combination and the object of the evaluation of the positive or negative, or neutral showing the Viewpoints of said fourth representation indicated by the notation, or the A rule that stores a combination of the object indicated by the first notation, the focus point indicated by the second notation, the positive, negative, or neutral evaluation indicated by the fourth notation and the property of the object in association with each other . Based on the property estimation means for estimating the property of the object corresponding to the partial text detected by the opinion expression location detection means,
A rule for storing the property indicated by the third notation and the focus of evaluation in association with each other when the second notation is missing in the partial text detected by the opinion expression location detecting unit, or A rule for storing a combination of an object indicated by the first notation and a property indicated by the third notation and a focus point of evaluation , or an object indicated by the first notation and the fourth Based on a rule that stores a combination of positive, negative, or neutral evaluations indicated by the notation and an evaluation focus in association with each other, the evaluation focus corresponding to the partial text detected by the opinion expression location detection means Focusing point estimation means for estimating
(A) the object corresponding to each opinion expression location, (b) the focus of evaluation regarding the object, (c) the nature of the object regarding the focus, (d) affirmation or denial or neutrality regarding the object An opinion extraction device comprising opinion element output means for outputting four items of evaluation. An opinion extraction method for detecting an opinion expression from an input text by an information processing device ,
A set of first notation that can indicate the object, a set of second notation that can indicate the focus of evaluation on the object, and a set of third notation that can indicate the properties of the object regarding the focus And a fourth set of notations that can indicate a positive, negative, or neutral evaluation regarding the object are registered in advance in the opinion element memory, and the first text registered in the opinion element memory from the input text. To an opinion element matching step for matching a character string corresponding to the fourth notation,
Based on the result of the opinion element collation step, an opinion statement location detecting step for detecting a partial text expressing an opinion from the input text in a state where the first to fourth notations are distinguished,
A rule for storing a property indicated by the third notation and an affirmative, negative, or neutral evaluation in association with each other when the fourth notation is missing in the partial text detected by the opinion expression location detecting step Or a rule for storing a combination of the focus point indicated by the second notation and the property indicated by the third notation and an evaluation of affirmation, negation, or neutrality , or the indication of the first notation The said opinion expression location detection step based on a rule that stores a combination of an object, a focus point indicated by the second notation, and a property indicated by the third notation, and an affirmative, negative, or neutral evaluation. And an evaluation estimation step of estimating an affirmative, negative, or neutral evaluation corresponding to the partial text detected by. An opinion extraction method for detecting an opinion expression from an input text by an information processing device ,
A set of first notation that can indicate the object, a set of second notation that can indicate the focus of evaluation on the object, and a set of third notation that can indicate the properties of the object regarding the focus And a fourth set of notations that can indicate a positive, negative, or neutral evaluation regarding the object are registered in advance in the opinion element memory, and the first text registered in the opinion element memory from the input text. To an opinion element matching step for matching a character string corresponding to the fourth notation,
Based on the result of the opinion element collation step, an opinion statement location detecting step for detecting a partial text expressing an opinion from the input text in a state where the first to fourth notations are distinguished,
When the third notation is missing in the partial text detected by the opinion expression location detecting step, the positive or negative or neutral evaluation indicated by the fourth notation is associated with the property of the object. storing rules, or the second rule in association with the nature of the combination and the object of the evaluation of the positive or negative, or neutral showing the Viewpoints of said fourth representation indicated by the notation, or the A rule that stores a combination of the object indicated by the first notation, the focus point indicated by the second notation, the positive, negative, or neutral evaluation indicated by the fourth notation and the property of the object in association with each other . And a property estimation step of estimating a property of the object corresponding to the partial text detected by the opinion expression location detection step. An opinion extraction method for detecting an opinion expression from an input text by an information processing device ,
A set of first notation that can indicate the object, a set of second notation that can indicate the focus of evaluation on the object, and a set of third notation that can indicate the properties of the object regarding the focus And a fourth set of notations that can indicate a positive, negative, or neutral evaluation regarding the object are registered in advance in the opinion element memory, and the first text registered in the opinion element memory from the input text. To an opinion element matching step for matching a character string corresponding to the fourth notation,
Based on the result of the opinion element collation step, an opinion statement location detecting step for detecting a partial text expressing an opinion from the input text in a state where the first to fourth notations are distinguished,
A rule for storing the property indicated by the third notation and the focus of evaluation in association with each other when the second notation is missing in the partial text detected by the opinion-expressing location detecting step, or A rule for storing a combination of an object indicated by the first notation and a property indicated by the third notation and a focus point of evaluation , or an object indicated by the first notation and the fourth Based on a rule that stores a combination of a positive or negative or neutral evaluation indicated by the notation and a focus of evaluation, and a focus of evaluation corresponding to the partial text detected by the opinion expression location detection step An opinion extraction method characterized by executing a point-of-interest estimation step for estimating. An opinion extraction method for detecting an opinion expression from an input text by an information processing device ,
A set of first notation that can indicate the object, a set of second notation that can indicate the focus of evaluation on the object, and a set of third notation that can indicate the properties of the object regarding the focus And a fourth set of notations that can indicate a positive, negative, or neutral evaluation regarding the object are registered in advance in the opinion element memory, and the first text registered in the opinion element memory from the input text. To an opinion element matching step for matching a character string corresponding to the fourth notation,
Based on the result of the opinion element collation step, an opinion statement location detecting step for detecting a partial text expressing an opinion from the input text in a state where the first to fourth notations are distinguished,
A rule for storing a property indicated by the third notation and an affirmative, negative, or neutral evaluation in association with each other when the fourth notation is missing in the partial text detected by the opinion expression location detecting step Or a rule for storing a combination of the focus point indicated by the second notation and the property indicated by the third notation and an evaluation of affirmation, negation, or neutrality , or the indication of the first notation The said opinion expression location detection step based on a rule that stores a combination of an object, a focus point indicated by the second notation, and a property indicated by the third notation, and an affirmative, negative, or neutral evaluation. A rating estimation step that estimates a positive, negative or neutral rating corresponding to the partial text detected by
When the third notation is missing in the partial text detected by the opinion expression location detecting step, the positive or negative or neutral evaluation indicated by the fourth notation is associated with the property of the object. storing rules, or the second rule in association with the nature of the combination and the object of the evaluation of the positive or negative, or neutral showing the Viewpoints of said fourth representation indicated by the notation, or the A rule that stores a combination of the object indicated by the first notation, the focus point indicated by the second notation, the positive, negative, or neutral evaluation indicated by the fourth notation and the property of the object in association with each other . Based on the property estimation step for estimating the property of the object corresponding to the partial text detected by the opinion expression location detection step,
A rule for storing the property indicated by the third notation and the focus of evaluation in association with each other when the second notation is missing in the partial text detected by the opinion-expressing location detecting step, or A rule for storing a combination of an object indicated by the first notation and a property indicated by the third notation and a focus point of evaluation , or an object indicated by the first notation and the fourth Based on a rule that stores a combination of a positive or negative or neutral evaluation indicated by the notation and a focus of evaluation, and a focus of evaluation corresponding to the partial text detected by the opinion expression location detection step A focus point estimation step for estimating
(A) the object corresponding to each opinion expression location, (b) the focus of evaluation regarding the object, (c) the nature of the object regarding the focus, (d) affirmation or denial or neutrality regarding the object An opinion extraction method characterized by executing an opinion element output step for outputting four items of evaluation. An opinion extraction program for causing an information processing device to detect an opinion expression part from input text,
The opinion extraction program uses the information processing apparatus,
A set of first notation that can indicate the object, a set of second notation that can indicate the focus of evaluation on the object, and a set of third notation that can indicate the properties of the object regarding the focus And an opinion element storage means in which is registered a fourth notation set that can indicate a positive, negative, or neutral evaluation of the object;
Opinion element matching means for matching a character string corresponding to the first to fourth notations registered in the opinion element storage means from the input text;
Based on the result of the opinion element matching means, a first composition pattern in which an opinion statement is composed of the first notation, the second notation, the third notation, and the fourth notation from the input text. An opinion statement is the second notation composed of the first notation, the second notation and the third notation, and an opinion statement is the first notation, the second notation and the fourth notation A third composition pattern composed of notations, a fourth composition pattern consisting of the first notation, the third notation and the fourth notation, and an opinion expression as the first notation and the first Opinion expression location detection means for detecting a partial text corresponding to any of the fifth composition pattern consisting of three notations and the sixth composition pattern whose opinion expression consists of the first notation and the fourth notation Opinion extraction professional characterized by functioning as Lamb. When the opinion expression location detection unit detects a partial text from the input text, the first pattern is set as a first priority, the second configuration pattern, the third configuration pattern, and the fourth configuration The second configuration pattern is set as the second priority order, and the fifth configuration pattern and the sixth configuration pattern are detected as the third priority order from the highest priority order. 16. Opinion extraction program according to 16. The opinion extraction program uses the information processing apparatus,
When the partial text detected by the opinion expression location detecting means is the second constituent pattern or the fifth constituent pattern, the property indicated by the third notation in the partial text and the positive / negative / neutral evaluation DOO or rules, in association with each other, combined with positive or negative, or neutral rating and the association with each of the properties listed and Viewpoints of said third representation indicated by the second representation in the partial text rule, or the evaluation of the combination with positive or negative, or neutral and properties listed with Viewpoints of said third representation indicated by the first of said second representation with the object indicated by the notation in partial text based on the rule that association with each other, evaluation estimation for estimating the evaluation of positive or negative or neutral corresponding to the detected partial text by the opinion portion detecting means Opinion extraction program according to claim 16 or claim 17 characterized in that to further function as a unit. The opinion extraction program uses the information processing apparatus,
When the partial text detected by the opinion expression location detection means is the third constituent pattern or the sixth constituent pattern, the positive and negative or neutral evaluation and target indicated by the fourth notation in the partial text Rules for storing the property of an object in association with each other , or a combination of a focus point indicated by the second notation in the partial text and an affirmation or negative or neutral evaluation indicated by the fourth notation and the property of the object preparative association with rule storing, or evaluation of positive or negative, or neutral shown Viewpoints showing the object of the second representation indicated by the first representation in the partial text and of the fourth representation and combined with based on the rules in association with the nature of the object, sex to estimate the properties of the object corresponding to the detected partial text by the opinion portion detecting means Opinion extraction program according to claim 18 claim 16 which is characterized in that to further function as estimating means. The opinion extraction program uses the information processing apparatus,
When the partial text detected by the opinion expression location detection means is the fourth composition pattern, the fifth composition pattern, or the sixth composition pattern, the property and evaluation indicated by the third notation in the partial text of Viewpoints and the correspondence with the rules stored or partial Viewpoints combination and evaluation of the first object indicated by the notation in the text and the nature indicated by the third representation and the association with each other Or a rule that stores a combination of the object indicated by the first notation and the positive or negative or neutral evaluation indicated by the fourth notation and the focus of evaluation in the partial text in association with each other . 17. The method according to claim 16, further functioning as a focus point estimation unit that estimates a focus point of evaluation corresponding to the partial text detected by the opinion expression location detection unit. Opinion extraction program according to any one of Motomeko 19. The opinion extraction program uses the information processing apparatus,
(A) the object corresponding to each opinion expression location, (b) the focus of evaluation regarding the object, (c) the nature of the object regarding the focus, (d) affirmation or denial or neutrality regarding the object The opinion extraction program according to any one of claims 16 to 20, further functioning as an opinion element output means for outputting four items, i.e. The opinion element output means outputs by distinguishing whether each item of (a), (b), (c), and (d) is described in the partial text or estimated. The opinion extraction program according to claim 21, wherein the opinion extraction program is made to function as follows. An opinion extraction device that detects an opinion expression from input text,
A set of first notation that can indicate the object, a set of second notation that can indicate the focus of evaluation on the object, and a set of third notation that can indicate the properties of the object regarding the focus And an opinion element storage means in which is registered a fourth notation set that can indicate a positive, negative, or neutral evaluation of the object;
Opinion element matching means for matching a character string corresponding to the first to fourth notations registered in the opinion element storage means from the input text;
Based on the result of the opinion element matching means, a first composition pattern in which an opinion statement is composed of the first notation, the second notation, the third notation, and the fourth notation from the input text. An opinion statement is the second notation composed of the first notation, the second notation and the third notation, and an opinion statement is the first notation, the second notation and the fourth notation A third composition pattern composed of notations, a fourth composition pattern consisting of the first notation, the third notation and the fourth notation, and an opinion expression as the first notation and the first Opinion expression location detection means for detecting a partial text corresponding to any of the fifth composition pattern consisting of three notations and the sixth composition pattern whose opinion expression consists of the first notation and the fourth notation An opinion extraction device characterized by comprising:

Images