JP2013131074A - Classification model learning method, device, program, and review document classifying method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To adjust a balance of positive examples and negative examples of learning data, and to accurately classify whether documents are review documents or not.SOLUTION: An evaluation sentence extracting portion 23 extracts evaluation sentences from each of a plurality of blog documents for learning which includes review documents, thereby producing evaluation documents. An identity extracting portion 24 extracts identities about each of the evaluation documents produced from each of the plurality of documents for learning. A learning portion 25 uses each of the extracted identities about the blog documents for learning which are the review documents as identities of positive examples, and each of the extracted identities about the blog documents for learning which are not the review documents as the identities of negative examples, thereby learning a classification model for learning whether the input documents are the review documents or not.

Description

  The present invention relates to a classification model learning method, apparatus, program, and review document classification method, and more particularly to a classification model learning method, apparatus, program, and review document classification for classifying whether or not a document is a review document. Regarding the method.

  There is a classification method for classifying a document using a word distribution in a document such as a blog (for example, Non-Patent Document 1). In this classification method, there are words that are likely to appear in a specific field and words that are not so, and a document is classified for each field by using their distribution.

Koichi Hirano, Kiya Kobayashi, Junichi Takahashi, "Automatic classification of Japanese-speaking blogs", IPSJ Research Report, 2005

  However, when a review document is extracted from a document obtained on social media such as a blog, the number of documents that can be a review document is smaller than that of a non-review document. For example, 80% of the blogs related to restaurants are non-reviewed documents, and there are only about 20% of review documents. Therefore, when using the learning device, the balance between the positive example and the negative example is deteriorated, and the classification performance of the classification model obtained as a learning result may be deteriorated.

  The present invention has been made in view of the above circumstances, and can perform classification model learning that can adjust the balance between positive and negative examples of learning data and can accurately classify whether or not the document is a review document. It is an object to provide a method, an apparatus, a program, and a review document classification method.

  In order to achieve the above object, the classification model learning method according to the first invention is determined in advance from each of a plurality of learning documents including a review document in which information about a specific object is described by an evaluation sentence extraction unit. Extracting the evaluation sentence including the evaluation expression, and the feature extraction means, for each of the plurality of learning documents from which the evaluation sentence is extracted by the evaluation sentence extraction means, the extracted from the learning document Each of the features extracted for the learning document that is the review document is corrected by the step of extracting the feature indicating the feature of the document composed of the evaluation sentence or the feature indicating the feature of the learning document. Each of the extracted features of the learning document that is not the review document is set as a negative example feature, and the input document is set as the feature of the example. Comprising the steps of: learning a classification model for classifying whether the-menu document, the.

  The classification model learning device according to the second invention is an evaluation sentence extraction means for extracting an evaluation sentence including a predetermined evaluation expression from each of a plurality of learning documents including a review document in which information on a specific object is described. And, for each of the plurality of learning documents from which the evaluation sentence has been extracted by the evaluation sentence extracting means, a feature indicating the characteristics of the document consisting of the evaluation sentence extracted from the learning document, or the learning document A feature extracting means for extracting a feature indicating a feature of the feature, and each of the features extracted for the learning document that is the review document is a positive feature, and the feature document is extracted for the learning document that is not the review document Learning means for learning a classification model for classifying whether or not the input document is the review document with each of the features as a negative feature. It is.

  According to the classification model learning method according to the first invention and the classification model learning device according to the second invention, each of the plurality of learning documents including the review document in which the information about the specific target is described by the evaluation sentence extraction unit. Then, an evaluation sentence including a predetermined evaluation expression is extracted. A feature indicating the characteristics of the document composed of the evaluation sentence extracted from the learning document, for each of the plurality of learning documents from which the evaluation sentence is extracted by the evaluation sentence extraction means by the feature extraction means, or A feature indicating the characteristics of the learning document is extracted.

  Then, each of the features extracted for the learning document that is the review document by the learning unit is set as a positive example feature, and each of the features extracted for the learning document that is not the review document is set as a negative example. As a feature, a classification model for classifying whether or not the input document is the review document is learned.

  As described above, by extracting an evaluation sentence from a learning document and learning a classification model based on each feature extracted for each of a plurality of learning documents from which the evaluation sentence has been extracted, The balance between the example and the negative example can be adjusted, and whether or not the document is a review document can be classified with high accuracy.

  A classification model learning method according to a third invention includes a step of dividing each of a plurality of learning documents including a review document in which information about a specific object is described by a sentence dividing unit, and an evaluation sentence extracting unit Extracting an evaluation sentence including a predetermined evaluation expression from each of the plurality of learning documents, and the plurality of learnings in which the evaluation sentence is extracted by the evaluation sentence extraction unit by a feature extraction unit. Extracting a feature indicating each feature of the evaluation sentence extracted from the learning document or a feature of each sentence of the learning document for each of the learning documents; The classification for classifying whether or not the inputted sentence is a sentence in the review document based on each evaluation sentence of the document or each of the features extracted for each sentence Including the steps of learning a Dell, a.

  In this way, each of the learning documents including the review document is divided into sentence units, the evaluation sentence is extracted from the learning document, and each feature extracted for each sentence of the learning document from which the evaluation sentence is extracted By learning the classification model based on the above, it is possible to adjust the balance between the positive example and the negative example of the learning data, and to classify whether or not the document is a review document with high accuracy.

  In the review document classification method according to the fourth aspect of the present invention, an input feature extraction unit extracts a feature indicating a feature of a document including the evaluation sentence in the input document or a feature indicating a feature of the input document. The input document is based on the classification model learned by the classification model learning method according to the first invention and the feature extracted by the input feature extraction means by the classification means; Categorizing whether the document is a review document.

  In this way, among the learning documents including the review document, the input document is the review document using the classification model learned based on each of the features extracted for the learning document from which the evaluation sentence is extracted. By determining whether or not the document is a review document, it is possible to accurately classify whether or not the document is a review document using learning data in which the balance between the positive example and the negative example is adjusted.

  A review document classification method according to a fifth aspect of the present invention includes a step of dividing an input document by a sentence unit by an input sentence dividing unit, and the evaluation sentence extracted from the input document by an input feature extracting unit. Each feature of the input document, or a step of extracting each sentence feature of the input document, the classification model learned by the classification model learning method according to the third invention by the classification means, and the input feature extraction means Classifying whether each evaluation sentence or each sentence of the input document is a sentence in the review document based on the feature of each evaluation sentence extracted by the above or the feature of each sentence The input document is converted to the review sentence based on the classification result of each evaluation sentence classified by the classification means or the classification result of each sentence of the document. Comprising the steps of: determining whether or not the.

  In this way, among the learning documents including the review document, the input document is reviewed using the classification model learned based on each feature extracted for each sentence of the learning document from which the evaluation sentence is extracted. By determining whether or not the document is a document, it is possible to accurately classify whether or not the document is a review document by using learning data in which the balance between the positive example and the negative example is adjusted.

  A program according to a sixth invention is a program for causing a computer to execute each step of the classification model learning method or the review document classification method.

  As described above, according to the classification model learning method, apparatus, and program of the present invention, the balance between positive and negative examples of learning data can be adjusted, and whether or not the document is a review document can be classified with high accuracy. The effect that it can do is acquired.

  Further, according to the review document classification method and program of the present invention, it is possible to accurately classify whether or not the document is a review document with learning data in which the balance between the positive example and the negative example is adjusted. .

It is the schematic which shows the structure of the review document classification | category apparatus which concerns on the 1st Embodiment of this invention. It is a figure which shows the blog document input. (A) The figure which shows the input blog document, (B) The figure which shows a morphological analysis result, (C) The figure which shows an evaluation document. It is a flowchart which shows the content of the learning process routine in the review document classification device which concerns on the 1st Embodiment of this invention. It is a flowchart which shows the content of the document classification | category process routine in the review document classification | category apparatus which concerns on the 1st Embodiment of this invention. It is the schematic which shows the structure of the review document classification | category apparatus based on the 2nd Embodiment of this invention. It is a figure for demonstrating the learning data obtained from a review document and a non-review document. It is a figure which shows the feature extracted about each evaluation sentence. It is a figure which shows the classification result with respect to each evaluation sentence. It is a flowchart which shows the content of the learning process routine in the review document classification device based on the 2nd Embodiment of this invention. It is a flowchart which shows the content of the document classification process routine in the review document classification device based on the 2nd Embodiment of this invention.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

[First embodiment]
<System configuration>
The review document classification apparatus 100 according to the first embodiment of the present invention receives a blog document obtained on social media (for example, a blog) and inputs objective information on a specific target (for example, a store or a product) or A determination result as to whether or not the document is a review document in which subjective information (for example, opinions such as word-of-mouth information) is described is output. One blog document is text data composed of one or more sentences. The review document classification apparatus 100 is configured by a computer including a CPU, a RAM, and a ROM that stores a program for executing a learning processing routine and a document classification processing routine described later. It is configured as follows. As shown in FIG. 1, the review document classification apparatus 100 includes an input unit 10, a calculation unit 20, and an output unit 30.

  The input unit 10 receives a document group including a plurality of blog documents input as learning documents. For example, data as shown in FIG. 2 can be input as a blog document. The input unit 10 receives, for each blog document, teacher information indicating whether the blog document is a review document related to a specific target, together with the blog document as a learning document.

  Further, the input unit 10 accepts a blog document input as a classification target.

  Note that the input blog document may have been subjected to morphological analysis. In this case, morphological analysis units 22 and 31 described later can be omitted.

  The calculation unit 20 includes a document database 21, a morpheme analysis unit 22, an evaluation sentence extraction unit 23, a feature extraction unit 24, a learning unit 25, and a model storage unit 26.

  The document database 21 stores a document group including a plurality of blog documents as learning documents received by the input unit 10 and teacher information for each blog document.

  For each blog document, the morpheme analysis unit 22 divides the sentence into words by morpheme analysis, which is an existing technology, and further gives a part of speech to each word for output. For example, if the blog document is “I bought a new product from Company A ...” as shown in FIG. 3 (A), the morphological analysis result is as shown in FIG. 3 (B). , “Company A (noun) / no (case particle) / new product (noun) / (case particle) / buy (verb stem) / tsu (verb inflection ending) / te (verb suffix) / shima (verb (Stem) / i (verb inflection ending) / ta (verb suffix) ... (omitted) ... ".

  The evaluation sentence extraction unit 23 extracts only evaluation sentences in which evaluation expressions prepared in advance appear for each blog document based on the morphological analysis result, and includes only evaluation sentences as shown in FIG. A set of evaluation documents composed only of evaluation sentences is created by creating the evaluated evaluation documents (documents from which non-evaluation sentences have been removed). At this time, a document in which an evaluation sentence does not appear is deleted from the set of evaluation documents.

Here, the evaluation sentences tend to appear more in the review document than in the non-review document.
By extracting non-evaluation sentences from the input learning blog document, it is possible to reduce the ratio of non-reviewed documents in the set of evaluation documents.

  The evaluation expression is acquired based on a dictionary of evaluation expressions created in advance. For example, “delicious”, “beautiful”, “beautiful”, “cute”, “fashionable”, “large”, “small”, “less”, “poor attitude”, “dirty” , “Bad”, etc. When the target is social media such as a blog, emoticons, pictograms, symbols, and the like may be used as evaluation expressions.

  For each blog document, the feature extraction unit 24 uses the result obtained by the morphological analysis for each created evaluation document to create a feature indicating the feature of the document used for machine learning. For example, the frequency distribution (histogram) of morphemes in the document is used as the document feature.

  There may be many specific emotional expressions and evaluation expressions appearing in review documents. Therefore, the presence / absence and type of sensitivity expression / evaluation expression may be used as a feature. In addition, emoticons and pictograms may be used as features.

  The feature extraction unit 24 stores the feature of the evaluation document created from the blog document as the review document in a memory (not shown) as positive example learning data. The feature extraction unit 24 stores the feature of the evaluation document created from the blog document, which is a non-review document, in the memory as negative example learning data.

  The learning unit 25 uses positive learning data (features of an evaluation document created from a review document) obtained from a document group as a learning document and negative learning data (an evaluation document created from a non-review document). A classification model for classifying whether or not the input document is a review document is created by machine learning using the feature, and stored in the model storage unit 26. As the machine learning algorithm, for example, an algorithm such as support vector machine (SVM) or Markov Logic Network (MLN) can be used.

  The classification model stored in the model storage unit 26 stores, for example, numerical values of weights regarding each feature.

  The computing unit 20 includes a morphological analysis unit 31, an evaluation sentence extraction unit 32, a feature extraction unit 33, and a classification unit 34. The feature extracting unit 33 is an example of an input feature extracting unit.

  Similar to the morpheme analysis unit 22, the morpheme analysis unit 31 divides the blog document to be classified into words by morpheme analysis, and further gives a part of speech to each word for output.

  Similar to the evaluation sentence extraction unit 32, the evaluation sentence extraction unit 32 extracts only evaluation sentences in which an evaluation expression prepared in advance appears for the blog document to be classified based on the morphological analysis result, and the evaluation document ( A document from which non-evaluation sentences are removed). At this time, if the blog document to be classified is a document in which an evaluation sentence does not appear, it may be classified as a non-reviewed document without performing subsequent processing.

  The feature extraction unit 33 uses the result obtained by the morphological analysis for the evaluation document created from the blog document to be classified, and creates a feature indicating the feature of the document, similar to the feature extraction unit 24.

  For example, for an evaluation document created from a blog document to be classified, for example, the classification unit 34 uses, for example, a feature vector whose elements are numerical values obtained by multiplying a numerical value indicating the extracted feature by a corresponding weight of the classification model. Is used to classify whether or not the document is a review document by the support vector machine algorithm.

  The classification result by the classification unit 34 is output from the output unit 30.

<Operation of review document classification device>
Next, the operation of the review document classification apparatus 100 according to the first embodiment will be described. First, when a document group which is a plurality of blog documents as learning documents and teacher information indicating whether each of the plurality of blog documents is a review document are input to the review document classification apparatus 100, a review is performed. The document grouping apparatus 100 stores the input document group and teacher information in the document database 21. Then, the review document classification apparatus 100 executes a learning process routine shown in FIG.

  First, in step S101, one blog document is extracted from the document database 21. In step S102, the morpheme analysis unit 22 performs a morpheme analysis process on the blog document extracted in step S101.

  In the next step S103, the evaluation sentence extraction unit 23 extracts an evaluation sentence from the blog document based on the morphological analysis result obtained in step S103 and a dictionary of evaluation expressions prepared in advance, and creates an evaluation document. To do. In step S104, the feature extraction unit 24 extracts features from the evaluation document created in step S103 based on the morphological analysis result obtained in step S102. In step S105, the blog document is a review document. If there is, the feature of the evaluation document extracted in step S104 is stored in the memory as the feature of the review document (learned data of the positive example). If the blog document is a non-review document, the feature is extracted in step S104. The characteristics of the evaluated document are stored in the memory as the characteristics of the non-reviewed document (negative learning data).

  In step S106, it is determined whether or not the processing in steps S101 to S105 has been executed for all the blog documents stored in the document database 21, and blog documents that have not executed the processing in steps S101 to S105 are determined. If it exists, the process returns to step S101 to take out the blog document. On the other hand, if the processes of steps S101 to S105 have been executed for all blog documents, the process proceeds to step S107.

  In step S107, the learning unit 25 learns the classification model by machine learning using the positive example learning data and the negative example learning data stored in the memory. In step S108, the classification model is stored in the model storage unit 26. Is stored, and the learning processing routine is terminated.

  When the blog document to be classified is input to the review document classification device 100, the review document classification device 100 executes a document classification processing routine shown in FIG.

  First, in step S111, a blog document input by the input unit 10 is received. In step S112, the morpheme analysis unit 31 performs morpheme analysis on the blog document input in step S111.

  In the next step S113, the evaluation sentence extraction unit 32 extracts an evaluation sentence from the blog document based on the morphological analysis result obtained in step S112 and a dictionary of evaluation expressions prepared in advance, and creates an evaluation document. To do.

  In the next step S114, the feature extraction unit 33 extracts features of the evaluation document of the blog document created in step S113. In step S115, the classification unit 34 classifies whether the blog document is a review document based on the features extracted in step S114 and the classification model stored in the model storage unit 26.

  In step S116, the classification result in step S115 is output by the output unit 30, and the document classification processing routine is terminated.

  As described above, according to the review document classification apparatus according to the first embodiment, an evaluation sentence is extracted from each of a plurality of learning blog documents, and the learning blog document including the evaluation sentence is included. The balance between the positive and negative examples of the learning data can be adjusted by learning the classification model based on each of the features extracted for each evaluation document created from It is possible to learn a classification model that can accurately classify these.

  Moreover, using the classification model learned based on each of the features extracted for the evaluation document created from the learning blog document that included the evaluation sentence among the learning blog documents including the review document, By determining whether or not the input blog document is a review document, it is possible to accurately classify whether or not it is a review document by using learning data in which the balance between the positive example and the negative example is adjusted.

[Second Embodiment]
<System configuration>
Next, a review document classification apparatus according to the second embodiment will be described. In addition, about the part which becomes the structure similar to 1st Embodiment, the same code | symbol is attached | subjected and description is abbreviate | omitted.

  The second embodiment is different from the first embodiment in that the blog document is divided into sentences, the features of each sentence are extracted, and the classification model is learned.

  As illustrated in FIG. 6, the calculation unit 220 of the review document classification apparatus 200 according to the second embodiment includes a document database 21, a sentence division unit 222, a morpheme analysis unit 22, an evaluation sentence extraction unit 23, and a feature extraction unit 224. The learning unit 225 and the model storage unit 26 are provided.

  The sentence division unit 222 divides each blog document of the document group stored in the document database 21 into sentence units. A known technique may be used as the sentence division method. For example, the sentence may be divided at a place where punctuation marks or line breaks appear. In addition, when using a document obtained on social media such as a blog, emoticons, pictograms, and the like are often used for sentence separation. Therefore, pictograms and emoticons may be used for sentence breaks.

  For each blog document, the feature extraction unit 224 creates a feature indicating the feature of the sentence, which is used for machine learning, using the result obtained by morphological analysis for each evaluation sentence of the created evaluation document. For example, a frequency distribution (histogram) of morphemes in a sentence is used as a sentence feature.

  Also, social media texts such as blogs may be evaluated across multiple sentences. For this reason, evaluation expressions of preceding and following sentences, morpheme frequencies, and the like may be used as features.

  As shown in FIG. 7, the feature extraction unit 224 stores the feature of each evaluation sentence extracted from the blog document as the review document in a memory (not shown) as positive example learning data. In addition, the feature extraction unit 224 stores the feature of each evaluation sentence extracted from the blog document, which is a non-review document, in the memory as negative example learning data.

  The learning unit 225 includes positive example learning data (features of evaluation sentences extracted from review documents) obtained from a document group that is a learning document and negative example learning data (of evaluation sentences extracted from non-review documents). A classification model for classifying whether or not the input sentence is a sentence in the review document is created by machine learning using the feature, and stored in the model storage unit 26.

  The calculation unit 220 includes a sentence division unit 231, a morpheme analysis unit 31, an evaluation sentence extraction unit 32, a feature extraction unit 233, a classification unit 234, and a review document determination unit 235. The sentence dividing unit 231 is an example of an input sentence dividing unit, and the feature extracting unit 233 is an example of an input feature extracting unit.

  Similar to the sentence dividing unit 222, the sentence dividing unit 231 divides the input blog document to be classified into sentence units.

  Similar to the morpheme analysis unit 22, the morpheme analysis unit 31 divides the sentence into words by morpheme analysis for each divided sentence of the blog document to be classified, and further gives a part of speech to each word for output.

  As shown in FIG. 8, the feature extraction unit 233 uses the result obtained by morphological analysis for each extracted evaluation sentence, as in the case of the feature extraction unit 224, as shown in FIG. Create features that show features.

  For example, for each evaluation sentence of the blog document to be classified, for example, the classification unit 234 uses a feature vector whose elements are each a numerical value obtained by multiplying a numerical value indicating the extracted feature by a corresponding weight of the classification model. Thus, it is classified by the support vector machine algorithm whether or not the evaluation sentence is a sentence in the review document. As a result, as shown in FIG. 9, each evaluation sentence is classified as a review sentence or a non-review sentence.

  The review document determination unit 235 determines that the blog document to be classified is a review document when the ratio of the evaluation sentence classified as the review sentence is equal to or higher than a threshold value. It is determined that the document is a non-review document.

  The determination result by the review document determination unit 235 is output from the output unit 30.

<Operation of review document classification device>
Next, the operation of the review document classification apparatus 200 according to the second embodiment will be described. In addition, about the process similar to 1st Embodiment, the same code | symbol is attached | subjected and detailed description is abbreviate | omitted.

  First, the learning process routine shown in FIG. 10 is executed by the review document classification apparatus 200.

  First, in step S101, one blog document is extracted from the document database 21. In step S102, the sentence dividing unit 222 divides the blog document extracted in step S101 into sentences. In step S102, the morpheme analysis unit 22 performs a morpheme analysis process on each sentence of the blog document extracted in step S101.

  In the next step S103, the evaluation sentence extraction unit 23 extracts an evaluation sentence from the blog document based on the morphological analysis result obtained in step S103 and a dictionary of evaluation expressions prepared in advance. In step S202, the feature extraction unit 224 extracts features for each of the evaluation sentences extracted in step S103 based on the morphological analysis result obtained in step S102. In step S203, the blog document is reviewed. If the document is a document, the feature of the evaluation sentence extracted in step S202 is stored in the memory as the feature of the review sentence (learned data of the positive example). If the blog document is a non-review document, the feature is read in step S202. The feature of the extracted evaluation sentence is stored in the memory as the feature of the non-review sentence (negative example learning data).

  In step S106, it is determined whether or not the processes in steps S101, S201, S102, S103, S202, and S203 have been executed for all blog documents stored in the document database 21, and the steps S101, S201, and S102 are performed. If there is a blog document that has not been subjected to the processes of S103, S202, and S203, the process returns to step S101, and the blog document is extracted. On the other hand, when the processes of steps S101, S201, S102, S103, S202, and S203 have been executed for all blog documents, the process proceeds to step S107.

  In step S107, the learning unit 225 learns the classification model by machine learning using the positive learning data and the negative learning data stored in the memory. In step S108, the classification model is stored in the model storage unit 26. Is stored, and the learning processing routine is terminated.

  When the blog document to be classified is input to the review document classification device 200, the review document classification device 200 executes a document classification processing routine shown in FIG.

  First, in step S111, a blog document input by the input unit 10 is received. In step S211, the sentence dividing unit 231 divides the blog document input in step S111 into sentence units. In step S112, the morpheme analysis unit 31 performs a morpheme analysis process on each sentence of the blog document input in step S111.

  In the next step S113, the evaluation sentence extraction unit 32 extracts an evaluation sentence from the blog document based on the morphological analysis result obtained in step S112 and a dictionary of evaluation expressions prepared in advance.

  In the next step S212, the feature extraction unit 233 extracts the features for each evaluation sentence of the blog document extracted in step S113. In step S213, for each evaluation sentence of the blog document by the classification unit 234, the evaluation sentence is converted into a review sentence based on the feature extracted in step S212 and the classification model stored in the model storage unit 26. Categorize whether it is a non-review sentence.

  In step S214, the review document determination unit 235 determines whether or not the blog document is a review document based on the ratio of the evaluation sentence classified as the review sentence in step S214. In step S215, the determination result in step S214 is output by the output unit 30, and the document classification processing routine is terminated.

  As described above, according to the review document classification apparatus according to the second embodiment, each of the learning blog documents including the review document is divided into sentence units, and the evaluation sentence is extracted from the learning blog document. By extracting and learning the classification model based on each feature extracted for each evaluation sentence of the learning blog document, the balance between the positive and negative examples of the learning data can be adjusted, and the review document It is possible to learn a classification model that can accurately classify whether or not.

  In addition, using a classification model learned based on each feature of each evaluation sentence extracted from a learning blog document including a review document, it is determined whether or not the input blog document is a review document. Thus, it is possible to accurately classify whether or not the document is a review document by using learning data in which the balance between the positive example and the negative example is adjusted.

  Note that the present invention is not limited to the above-described embodiment, and various modifications and applications are possible without departing from the gist of the present invention.

  For example, a non-evaluation sentence may not be removed from a learning blog document including an evaluation sentence. That is, the learning blog document that does not include the evaluation sentence may be simply deleted from the set of learning blog documents without creating the evaluation document obtained by removing the non-evaluation sentence from the learning blog document. When the document is not divided into sentences as in the first embodiment described above, the classification feature may be learned by extracting the document features of the blog document including the evaluation sentence. Further, the document features of the blog document to be classified may be extracted, and the classification model may be used to classify whether or not the blog document is a review document. When the document is divided into sentences as in the second embodiment, the feature of each sentence (evaluation sentence, non-evaluation sentence) is extracted from the blog document including the evaluation sentence, and the classification model is extracted. Just learn. In addition, the feature of each sentence is extracted for the blog document to be classified, and the classification model is used to classify whether each sentence of the blog document is a review sentence. What is necessary is just to determine whether the said blog document is a review document based on a ratio.

  In the second embodiment, each evaluation sentence of the blog document for learning is a review sentence in which objective information or subjective information about the specific object is described, or a non-review sentence. You may make it input manually whether there exists. In this case, each feature extracted for each review sentence in each evaluation sentence of the learning blog document is used as positive learning data, and each non-review sentence in each sentence of the learning blog document is used. The classification model may be learned by the learning unit using each of the features extracted for as learning data of a negative example. If non-evaluation sentences are not removed from learning blog documents that include evaluation sentences, each sentence of learning blog documents that include evaluation sentences is either a review sentence or a non-review sentence. You may make it input manually.

  In the present specification, the embodiment has been described in which the program is installed in advance. However, the program can be provided by being stored in a computer-readable recording medium.

DESCRIPTION OF SYMBOLS 10 Input part 20, 220 Operation part 21 Document database 22, 31 Morphological analysis part 23, 32 Evaluation sentence extraction part 24, 33, 224, 233 Feature extraction part 25, 225 Learning part 26 Model storage part 34, 234 Classification part 100, 200 Review Document Classification Device 222, 231 Sentence Division Unit 235 Review Document Determination Unit

Claims (8)

A step of extracting an evaluation sentence including a predetermined evaluation expression from each of a plurality of learning documents including a review document in which information on a specific object is described by an evaluation sentence extraction unit;
A feature indicating the characteristics of the document composed of the evaluation sentence extracted from the learning document, for each of the plurality of learning documents from which the evaluation sentence is extracted by the evaluation sentence extraction means by the feature extraction means, or Extracting features indicating characteristics of the learning document;
Each of the features extracted for the learning document that is the review document by the learning means is a positive feature, and each of the features extracted for the learning document that is not the review document is a negative feature. Learning a classification model for classifying whether or not the input document is the review document;
Classification model learning method including Dividing each of a plurality of learning documents including a review document in which information on a specific object is described by a sentence dividing unit;
A step of extracting an evaluation sentence including a predetermined evaluation expression from each of the plurality of learning documents by an evaluation sentence extracting unit;
A feature indicating the characteristics of each of the evaluation sentences extracted from the learning document for each of the plurality of learning documents from which the evaluation sentence has been extracted by the evaluation sentence extraction means by the feature extraction means, or the learning Extracting features of each sentence of the document for use,
For classifying whether or not the input sentence is a sentence in the review document based on each evaluation sentence of each of the plurality of learning documents or each of the features extracted for each sentence by the learning means. Learning the classification model;
Classification model learning method including   The step of learning by the learning means includes setting each of the features extracted for each evaluation sentence or each sentence of the learning document that is the review document as a positive example feature, and each of the learning documents that are not the review document. The classification model learning method according to claim 2, wherein the classification model is learned by using each of the features extracted for the evaluation sentence or each sentence as a negative example feature.   The step of learning by the learning means sets each of the features extracted for each of the evaluation sentences of the learning document or each of the review sentences in which information on the specific target of the sentences is described as a positive example feature. 3. The classification model according to claim 2, wherein the classification model is learned by using each of the extracted features for each evaluation sentence or each sentence of the learning document that is not the review sentence as a negative feature. Learning method. Extracting a feature indicating a feature of the document composed of the evaluation sentence in the input document or a feature indicating the feature of the input document by an input feature extracting unit;
Whether the input document is the review document based on the classification model learned by the classification model learning method according to claim 1 and the feature extracted by the input feature extraction means. Classifying whether or not,
Review document classification method including Dividing the input document into sentence units by the input sentence dividing means;
Extracting each feature of the evaluation sentence extracted from the input document or a feature of each sentence of the input document by an input feature extraction unit;
The classification model learned by the classification model learning method according to any one of claims 2 to 4, and the features of each evaluation sentence extracted by the input feature extraction means or each sentence Categorizing whether or not each evaluation sentence or each sentence of the input document is a sentence in the review document based on the feature;
Determining whether or not the input document is the review document based on a classification result of each evaluation sentence classified by the classification means or a classification result of each sentence of the document by the determination means;
Review document classification method including An evaluation sentence extracting means for extracting an evaluation sentence including a predetermined evaluation expression from each of a plurality of learning documents including a review document in which information on a specific object is described;
For each of the plurality of learning documents from which the evaluation sentence has been extracted by the evaluation sentence extracting means, a feature indicating the characteristics of the document composed of the evaluation sentence extracted from the learning document, or the characteristics of the learning document A feature extracting means for extracting a feature indicating
Each of the features extracted for the learning document that is the review document is input as a positive example feature, and each of the features extracted for the learning document that is not the review document is input as a negative example feature Learning means for learning a classification model for classifying whether or not a document is the review document;
Model learning apparatus including   The program for making a computer perform each step of the classification model learning method of any one of Claims 1-4, or the review document classification method of Claim 5 or 6.