JP3955522B2 - Data analysis apparatus and method, and program - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a data analysis apparatus and method using a text mining technique, and a program.
[0002]
[Prior art]
In recent years, in order to effectively formulate a management strategy or a marketing strategy, it is indispensable to derive a trend or pattern up to the present time necessary for management or marketing and perform data analysis. Data mining is applied as a technique for deriving trends and patterns in this data analysis. When handling customer information in data mining, the analysis is based on quantitative customer information (attribute information) based on unambiguous and specific facts based on questionnaire results.
[0003]
For example, as data mining from purchase history, numerical information such as customer age, purchase date and time, attribute data related to purchased items is used as input data, and the causal relationship of attribute data is mined to perform cross-sell / up-sell analysis. Is what you do. As an example of the application of data mining in product development situations, taking notebook PC development as an example, the perceptual quality is evaluated by methods such as conjoint analysis from the combination conditions of attribute data such as battery operating time and the presence or absence of DVD drive. It is.
[0004]
In analysis in data mining, attribute information based on such unambiguous and specific facts is analyzed, and the customer's subjective intention behind the analysis is not analyzed. For this reason, for example, (1) the relationship of the place, time, situation, etc. where the customer has evaluated is not considered. (2) I don't know what kind of thought the customer came to purchase behavior. (3) I do not know why the customer purchased the product in that situation. (4) I don't know what kind of idea the particular model was evaluated high (low). There was a problem such as.
[0005]
Therefore, text mining that proposes not only analyzing unambiguous and specific information as in the above-described data mining but also analyzing the result of a customer answering in text has been proposed. The customer's response sentence and the like targeted by text mining may include the customer's subjective intention, and it is expected that more useful information for management and marketing can be obtained.
[0006]
[Problems to be solved by the invention]
However, many text mining techniques up to now have followed the data mining technique. Here, the words contained in the text are regarded as keywords, as unambiguous and specific data, and are analyzed only in the same way as data mining, and the customer's subjective intention behind the words There was almost no analysis.
[0007]
For this reason, the information regarding the customer's subjective intention was missing from the data for management analysis or marketing analysis generated as a result. In addition, text mining that only follows the data mining method has the same quality as the data mining result, and it is necessary to formulate management strategy and marketing strategy from the resulting data. There were many parts that had to rely on the experience and intuition of strategy makers.
[0008]
It is an object of the present invention to provide a data analysis apparatus and method, and a program capable of accurately analyzing data by capturing subjective information using a text mining technique.
[0009]
It is another object of the present invention to provide a data analysis apparatus and method, and a program that can provide an analysis result that facilitates the formulation of a management strategy, a marketing strategy, and the like.
[0010]
[Means for Solving the Problems]
  In order to achieve the above object, a data analysis apparatus according to the first aspect of the present invention includes:
  A noun phrase dictionary that registers noun phrases that can be included in sentences to be analyzedNoun phrase dictionary storage means for storingWhen,
  A modifier phrase dictionary registered by associating modifier phrases that can modify noun phrases registered in the noun phrase dictionary with level values indicating whether each modifier phrase is semantically preferable or not preferredModifier dictionary storage means for storingWhen,
By instructing the start of sentence analysis,SaidStored in the noun phrase dictionary storage meansRefer to the noun phrase dictionaryIncludes a description part about the current state and a description part about the future state related to the current stateTo be analyzedEach description part included in the sentenceNoun phrase extraction means for extracting noun phrases from
  SaidStored in the modifier dictionary storage meansReferring to a modifier phrase dictionary, a modifier phrase extracting means for extracting a modifier phrase that modifies the noun phrase extracted by the noun phrase extracting means;
  Among the sentences to be analyzed, the tense used is the past tense, present tense or future tense, the described position is first or second, or registered in the dictionary stored in the storage means A separation means for separating a description part related to the current state and a description part related to the future state according to whether or not a frequent phrase of the future state is included;
For each description part carved by the carving means,The level value registered in the modifier phrase dictionary for the modifier phrase extracted by the modifier phrase extraction unit is totaled for each noun phrase extracted by the noun phrase extraction unit, and the total level value is calculated for each noun phrase. A degree of effect calculation means for calculating as
  An effect degree output means for outputting the effect degree for each noun phrase calculated by the effect degree calculation means;
  It is characterized by providing.
[0011]
  The data analysis apparatus according to the first aspect is as follows.
  For each description part carved by the carving means,It becomes the analysis object of each noun phrase extracted by the noun phrase extracting meansEach description part included in the sentenceThere may be further provided a noun phrase frequency calculating means for calculating the appearance frequency in. in this case,
  The effect level output means may output the degree of effect calculated by the effect level calculation means for each noun phrase and the appearance frequency calculated by the noun phrase frequency calculation means in a visual association.
[0012]
  The data analysis apparatus according to the first aspect is as follows.
  SaidStored in the modifier dictionary storage meansRefer to the modifier dictionary for analysisEach description part included in the sentenceA modifier extraction means for extracting a modifier that modifies the noun phrase in
  SaidStored in the noun phrase dictionary storage meansA noun phrase extraction unit that refers to a noun phrase dictionary and extracts a noun phrase that is modified by each of the modifiers extracted by the modifier extraction unit;
  For each description part carved by the carving means,Based on the extraction result of the noun phrase extracted by the noun phrase extracting means, the degree of aggregation is calculated for each modifier phrase indicating the degree to which the modifier phrase extracted by the modifier extraction means modifies the same noun phrase. Means,
  An aggregation level output unit that outputs an aggregation level for each modifier phrase calculated by the aggregation level calculation unit may be further provided.
[0013]
  In this case, the data analysis apparatus according to the first aspect is
  For each description part carved by the carving means,It becomes the analysis object of each modifier phrase extracted by the modifier phrase extraction meansEach description part included in the sentenceThere may be further provided a modifier phrase frequency calculating means for calculating the appearance frequency in. And
  The degree-of-aggregation output means may output the degree-of-aggregation calculated by the degree-of-aggregation calculation means for each modifier and the appearance frequency calculated by the modifier phrase frequency calculation means in a visual association.
[0014]
  In order to achieve the above object, a data analysis apparatus according to the second aspect of the present invention provides:
  A noun phrase dictionary that registers noun phrases that can be included in sentences to be analyzedNoun phrase dictionary storage means for storingWhen,
  A modifier phrase dictionary that registers modifier phrases that can modify noun phrases registered in the noun phrase dictionaryModifier dictionary storage means for storingWhen,
By instructing the start of sentence analysis,SaidStored in the modifier dictionary storage meansRefer to the modifier dictionaryIncludes a description part about the current state and a description part about the future state related to the current stateTo be analyzedEach description part included in the sentenceA modifier extraction means for extracting a modifier that modifies the noun phrase in
  SaidStored in the noun phrase dictionary storage meansA noun phrase extraction unit that refers to a noun phrase dictionary and extracts a noun phrase that is modified by each of the modifiers extracted by the modifier extraction unit;
  Among the sentences to be analyzed, the tense used is the past tense, present tense or future tense, the described position is first or second, or registered in the dictionary stored in the storage means A separation means for separating a description part related to the current state and a description part related to the future state according to whether or not a frequent phrase of the future state is included;
For each description part carved by the carving means,Based on the extraction result of the noun phrase extracted by the noun phrase extracting means, the degree of aggregation is calculated for each modifier phrase indicating the degree to which the modifier phrase extracted by the modifier extraction means modifies the same noun phrase. Means,
  An intensity output means for outputting the intensity for each modifier calculated by the intensity calculation means;
  It is characterized by providing.
[0015]
  The data analysis apparatus according to the second aspect is as follows.
  For each description part carved by the carving means,It becomes the analysis object of each modifier phrase extracted by the modifier phrase extraction meansEach description part included in the sentenceThere may be further provided a modifier phrase frequency calculating means for calculating the appearance frequency in. In this case,
  The degree-of-aggregation output means may output the degree-of-aggregation calculated by the degree-of-aggregation calculation means for each modifier and the appearance frequency calculated by the modifier phrase frequency calculation means in a visual association.
[0018]
  In order to achieve the above object, a data analysis method according to the third aspect of the present invention includes:
  A noun phrase dictionary that registers noun phrases that can be included in the sentence to be analyzed, and a modifier that can modify the noun phrases registered in the noun phrase dictionary, whether each modifier is semantically preferable or not preferable The modifier dictionary registered in association with the level valueIn addition to storing dictionary storage means, it has storage means used as a work areaIn a computer device,
By instructing the start of analysis,SaidStored in dictionary storage meansRefer to the noun phrase dictionaryStored in the storage means, including a description part relating to the current state and a description part relating to the future state relating to the current stateTo be analyzedEach description part included in the sentenceExtract noun phrases from, Storing the extracted noun phrase in the storage means;
  SaidStored in dictionary storage meansRefer to the modifier dictionary and extractStored in the storage meansExtract modifiers that modify noun phrases, Storing the extracted modifier in the storage means;
Among the sentences to be analyzed stored in the storage means, the tense used is the past tense, the present tense or the future tense, the described position is the first or the last, or the dictionary storage means According to whether or not to include a frequent phrase of the future state registered in the stored frequent phrase dictionary, the description part regarding the current state and the description part regarding the future state are separated and stored in the storage means,
  For each separated description part,ExtractStored in the storage meansThe level value registered in the modifier dictionary for the modifier is extracted.Stored in the storage meansAggregate for each noun phrase, and calculate the level value for each noun phrase as the degree of effect for each noun phrase., Storing the calculated degree of effect in the storage means,
  CalculateStored in the storage meansThe degree of effect for each noun phraseFrom the output device of the computer deviceoutputMake
  It is characterized by that.
[0020]
  In order to achieve the above object, a data analysis method according to the fourth aspect of the present invention includes:
  A noun phrase dictionary in which noun phrases that can be included in a sentence to be analyzed are registered, and a modifier phrase dictionary in which modifier phrases that can modify noun phrases registered in the noun phrase dictionary are registered.In addition to storing dictionary storage means, it has storage means used as a work areaIn a computer device,
  By instructing the start of analysis,SaidStored in dictionary storage meansRefer to the modifier dictionaryStored in the storage means, including a description part relating to the current state and a description part relating to the future state relating to the current stateTo be analyzedEach description part included in the sentenceExtract modifiers that modify noun phrases in, Storing the extracted modifier in the storage means;
  SaidStored in dictionary storage meansRefer to the noun phrase dictionary and extractStored in the storage meansExtract noun phrases modified by each modifier,
Among the sentences to be analyzed stored in the storage means, the tense used is the past tense, the present tense or the future tense, the described position is the first or the last, or the dictionary storage means According to whether or not to include a frequent phrase of the future state registered in the stored frequent phrase dictionary, the description part regarding the current state and the description part regarding the future state are separated and stored in the storage means,
  For each separated description part,Based on the extraction result of the noun phrase, the extraction is performed.Stored in the storage meansThe degree of aggregation that indicates the degree to which the modifiers modify the same noun phrase is calculated for each modifier., Storing the calculated degree of aggregation in the storage means,
  CalculateStored in the storage meansThe degree of aggregation for each modifierFrom the output device of the computer deviceoutputMake
  A data analysis method characterized by that.
[0021]
The data analysis method according to the fourth aspect is as follows.
The appearance frequency of the extracted modifiers in the sentence to be analyzed can be further increased. In this case,
The degree of aggregation may be visually output for each modifier phrase in a visual association with the calculated appearance frequency.
[0022]
  In order to achieve the above object, a program according to the fifth aspect of the present invention is:
  A noun phrase dictionary that registers noun phrases that can be included in the sentence to be analyzed, and a modifier that can modify the noun phrases registered in the noun phrase dictionary, whether each modifier is semantically preferable or not preferable The modifier dictionary registered in association with the level valueA stored dictionary storage means is provided.Computer equipment,
By instructing the start of sentence analysis,SaidStored in dictionary storage meansRefer to the noun phrase dictionaryIncludes a description part about the current state and a description part about the future state related to the current stateTo be analyzedEach description part included in the sentenceNoun phrase extraction means for extracting noun phrases from
  SaidStored in dictionary storage meansA modifier extraction unit that refers to a modifier dictionary and extracts a modifier that modifies the noun phrase extracted by the noun phrase extraction unit;
  Among the sentences to be analyzed, the tense used is the past tense, present tense or future tense, the described position is first or second, or registered in the dictionary stored in the storage means A separation means for separating a description part about the current state and a description part about the future state according to whether or not a frequent phrase of the future state is included,
For each description part carved by the carving means,The level value registered in the modifier phrase dictionary for the modifier phrase extracted by the modifier phrase extraction unit is totaled for each noun phrase extracted by the noun phrase extraction unit, and the total level value is calculated for each noun phrase. A degree-of-effect calculation means for calculating
  The degree of effect for each noun phrase calculated by the degree of effect calculating means isOutput from an output device included in the computer deviceAbility degree output means and
  It is made to function as.
[0024]
  In order to achieve the above object, a program according to the sixth aspect of the present invention is:
  A noun phrase dictionary in which noun phrases that can be included in a sentence to be analyzed are registered, and a modifier phrase dictionary in which modifier phrases that can modify noun phrases registered in the noun phrase dictionary are registered.A stored dictionary storage means is provided.Computer equipment,
  A noun phrase dictionary that registers noun phrases that can be included in sentences to be analyzedNoun phrase dictionary storage means for storing,
  A modifier phrase dictionary that registers modifier phrases that can modify noun phrases registered in the noun phrase dictionaryModifier dictionary storage means for storing,
By instructing the start of sentence analysis,SaidStored in the modifier dictionary storage meansRefer to the modifier dictionaryIncludes a description part about the current state and a description part about the future state related to the current stateTo be analyzedEach description part included in the sentenceA modifier extraction means for extracting a modifier that modifies the noun phrase in
  SaidStored in the noun phrase dictionary storage meansA noun phrase extraction unit that refers to a noun phrase dictionary and extracts a noun phrase that is modified by each modifier extracted by the modifier extraction unit;
  Among the sentences to be analyzed, the tense used is the past tense, present tense or future tense, the described position is first or second, or registered in the dictionary stored in the storage means A separation means for separating a description part about the current state and a description part about the future state according to whether or not a frequent phrase of the future state is included,
For each description part carved by the carving means,Based on the extraction result of the noun phrase extracted by the noun phrase extracting means, the degree of aggregation is calculated for each modifier phrase indicating the degree to which the modifier phrase extracted by the modifier extraction means modifies the same noun phrase. Means and
  The degree of aggregation for each modifier calculated by the degree of aggregation calculation meansOutput from an output device included in the computer deviceAggregation level output means
  It is made to function as.
[0025]
The program according to the sixth aspect provides the computer device,
You may make it function further as a modifier phrase frequency calculation means which calculates the appearance frequency in the sentence used as the analysis object of each modifier phrase extracted by the modifier phrase extraction means. In this case,
The degree-of-aggregation output means may output the degree-of-aggregation calculated by the degree-of-aggregation calculation means for each modifier and the appearance frequency calculated by the modifier phrase frequency calculation means in a visual association.
[0026]
DETAILED DESCRIPTION OF THE INVENTION
Embodiments of the present invention will be described below with reference to the accompanying drawings.
[0027]
FIG. 1 is a block diagram showing a system configuration centering on a data analysis apparatus according to this embodiment. The data analysis device 1 is connected to a terminal device (such as a personal computer or a workstation) 2 used by employees in a management department via a LAN (Local Area Network) 3. Further, it is also connected to a customer terminal device (such as a personal computer or a mobile phone) 4 from the gateway 5 via the Internet 6. The data analysis device 1 includes a CPU (Central Processing Unit) 11, a storage device 12, a communication device 13, and a file device 14.
[0028]
The CPU 11 executes a program stored in the storage device 12 and performs data analysis by a text mining technique as will be described later. The storage device 12 includes a main storage device and an auxiliary storage device, stores a program executed by the CPU 11, and is used as a work area for the CPU 11. The communication device 13 transmits / receives information to / from the terminal devices 2 and 4 via the LAN 3 and the Internet 6.
[0029]
The file device 14 includes a questionnaire file 14a, a knowledge dictionary 14b, and an analysis result database 14c. The file device 14 is physically included in the auxiliary storage device of the storage device 12, but includes a file and a database having an important role in the present invention. Yes.
[0030]
The questionnaire file 14a is a file in which text data of a questionnaire result input by a customer in a natural language sentence for a product is recorded. This questionnaire is entered in the terminal device 4 according to the entry form 100 as shown in FIG. 2, transmitted to the data analysis apparatus 1 via the Internet 6, and recorded in the questionnaire file 14a.
[0031]
The entry form 100 in FIG. 2 is an entry column (satisfaction entry column) 101 of points that the customer feels convenient for the products targeted by the questionnaire and what to expect to further extend the useful points. Field (progress expectation entry field) 102, entry field of points that the customer feels inconvenient (dissatisfaction entry field) 103, and entry field of how to improve the inconvenience point (improvement expectation entry field) 104. Here, the entry columns 101 and 103 are filled with the current state, and the entry columns 102 and 104 are filled with the future state.
[0032]
The knowledge dictionary 14b includes a word dictionary, a grammar dictionary, and a part of speech dictionary. As the word dictionary and the grammar dictionary, substantially the same dictionary as conventionally used for morphological analysis and syntactic analysis of sentences is applied. The part-of-speech dictionary does not register words with grammatical meanings registered in the word dictionary, but rather the words to be evaluated (single words, phrases that have a single meaning by concatenating multiple words) Is registered. The part-of-speech dictionary is prepared separately for nouns, adjectives and verbs, but these are not classified according to grammatical functions, but they function as nouns in meaning or as adjectives. Or whether it works verbally.
[0033]
The noun phrase dictionary is a dictionary in which noun phrases modified by adjective phrases and verb phrases are registered. The adjective phrase dictionary and the verb phrase dictionary are dictionaries that respectively register adjective phrases and verb phrases that will modify the noun phrase. The modification / modification relationship here does not require grammatically whether or not it becomes a modification / modification word, but may have a dependency relationship semantically. Moreover, it does not ask | require whether the modifier is preceded or followed by the word to be modified. In particular, the adjective phrase dictionary is associated with adjective phrases (expression words and normalized expressions) as shown in FIG. 3 in order to perform the evaluation analysis described later, and whether the word evaluates preferably or not. The level value shown is registered.
[0034]
The analysis result database 14c is a database for registering the results of data analysis described later with reference to the knowledge dictionary 14b for the questionnaire text recorded in the questionnaire file 14a. The analysis result database 14c may also register the analysis / analysis results (the results of morpheme analysis and syntactic analysis, which will be described later, and the results of subject analysis) before being output as the final results of data analysis.
[0035]
Hereinafter, processing in the data analysis apparatus 1 according to this embodiment will be described. Here, it is assumed that the questionnaire to be subjected to text mining has already been sent from the terminal device 4 and recorded in the questionnaire file 14a, but has not been analyzed yet. FIG. 4 is a flowchart showing processing in the data analysis apparatus 1. The process of this flowchart is started when an instruction to start processing is sent from the terminal device 2.
[0036]
When the processing starts, the CPU 11 first refers to the word dictionary and grammar dictionary of the knowledge dictionary 14b, morphologically analyzes the sentences of each questionnaire recorded in the questionnaire file 14a (step S11), and performs syntax analysis (step S11). S12). The morpheme analysis and the syntax analysis are performed by the same method as the conventional method, and it becomes possible to grasp a modification / modification relationship to be described later based on the result of the syntax analysis.
[0037]
When the morphological analysis and syntax analysis are completed, the CPU 11 next performs the subject analysis (step S13). In the subject analysis, the CPU 11 divides each questionnaire recorded in the questionnaire file 14a into sub-document units, refers to the noun phrase dictionary in the knowledge dictionary 14b, and extracts a noun phrase from each sub-document. CPU11 calculates | requires statistical information, such as the appearance frequency in a subdocument, the distribution in the whole questionnaire, etc. with respect to each extracted noun phrase, and converts each subdocument into a vector expression using this statistical information.
[0038]
CPU11 produces | generates the vector expression for every questionnaire based on the vector expression of each subdocument, calculates | requires the similarity of the vector expression for every questionnaire, and makes the documents of the questionnaire exceeding a predetermined value into one cluster. CPU11 calculates | requires the similarity of the vector expression of each cluster, and collects the clusters exceeding a predetermined value as the same subject. One questionnaire may be classified into a plurality of subjects.
[0039]
When the subject analysis ends, the CPU 11 performs data analysis on the subject obtained by the subject analysis, and extracts problems and the like described in the questionnaire (step S14). In the data analysis, evaluation analysis (step S14-1), sensibility analysis (step S14-2), and function requirement analysis (step S14-3) are performed based on a viewpoint that becomes a starting point when extracting problems and the like. .
[0040]
This data analysis is divided into a description relating to the current state (items entered in the entry fields 101 and 103 of the entry form 100) and a description relating to the future state (items entered in the entry fields 102 and 104 of the entry form 100). It can be carried out. Here, the subjects classified by the subject analysis in step S13 are displayed in a simple dendrogram as shown in FIG. 5, for example, and the operator selects a subject for which data analysis is to be performed.
[0041]
In the evaluation analysis in step S14-1, the CPU 11 refers to the noun phrase dictionary, extracts noun phrases included in the selected subject classification, and obtains the appearance frequency (frequency) for each noun phrase. The CPU 11 further extracts an adjective phrase that modifies each noun phrase by referring to the parsing result and the adjective phrase dictionary, and acquires the degree of affect of each noun phrase according to the level value of the extracted adjective phrase. The acquisition of the degree of effect will be described in more detail. When an adjective phrase that modifies a noun phrase is extracted, a level value corresponding to the adjective phrase registered in the adjective phrase dictionary in the knowledge dictionary 14b is acquired. The result of summing up the level values for each noun phrase according to the following formula 1 (which may be negatively calculated) is the degree of effect of each noun phrase.
[0042]
[Expression 1]
Affectiveness = Σ ((level value of adjective phrase) × (frequency of phrase consisting of modified noun phrase and adjective phrase)) ÷ (frequency of phrase including modified noun phrase)
However, Σ is the sum of the weight level calculation results for each adjective phrase categorized by reducing the modified noun phrase
[0043]
Note that reducing means summing up / aggregating common / overlapping parts based on conditions specified among a plurality of processing targets. As a result of the aggregation, the common / overlapping part is unique (reduced), and the common / non-overlapping part is categorized as difference information. For example, for the three phrases (adjective phrase-noun phrase) of "beautiful-color", "beautiful-color", and "nice-color", the modified noun phrase is reduced to "color-beautiful / beautiful / beautiful". "
[0044]
In the sentiment analysis in step S14-2, the CPU 11 refers to the adjective phrase dictionary, extracts adjective phrases included in the selected subject classification, and obtains the appearance frequency (frequency) of each adjective phrase. The CPU 11 further refers to the syntax analysis result and the noun phrase dictionary, and acquires the evaluation aggregation degree indicating how fixed the noun phrase that each adjective phrase modifies. Here, the evaluation intensity is obtained according to the following formula 2. If the variation of the noun phrase in the modification / modification relationship with the same adjective phrase is less than the frequency of the adjective phrase, the evaluation intensity of the adjective phrase Becomes higher.
[0045]
[Expression 2]
Evaluation intensity = -log (number of different noun phrases / frequency of adjective phrases)
However, only noun phrases and adjective phrases that are in a modified / modified relationship are counted.
[0046]
In the function request analysis in step S14-3, the CPU 11 refers to the verb phrase dictionary, extracts verb phrases that modify noun phrases included in the selected subject classification, and the appearance frequency of each verb phrase ( Frequency). The CPU 11 further refers to the parsing result and the noun phrase dictionary, and acquires the opinion aggregation degree indicating how fixed the noun phrase that each verb phrase modifies. Here, the opinion aggregation level is obtained according to the following Equation 3. If the variation of the noun phrase in the modification / modification relationship with the same verb phrase is less than the frequency of the verb phrase, the opinion aggregation level of the verb phrase Becomes higher.
[0047]
[Equation 3]
Opinion intensity = -log (number of different noun phrases / frequency of verb phrases)
However, only noun phrases and verb phrases that are in a modified / modified relationship are counted.
[0048]
The extraction of noun phrases, adjective phrases, or verb phrases in the evaluation analysis in step S14-1, the sensitivity analysis in step S14-2, and the function requirement analysis in step S14-3 will be described in detail later.
[0049]
When the data analysis for the selected subject is completed, the CPU 11 registers the analysis result in the analysis result database 14c (step S15). Further, the CPU 11 causes the communication device 13 to transmit the analysis result to the terminal device 2 that has instructed to start processing, and displays and outputs the analysis result on the display device of the terminal device 2 (step S16). This completes the processing in the data analysis device 1. The analysis result registered in the analysis result database 14c can be taken out at any time by making a request from the terminal device 2.
[0050]
Next, extraction of noun phrases, adjective phrases, or verb phrases in the evaluation analysis in step S14-1, the sensitivity analysis in step S14-2, and the function requirement analysis in step S14-3 will be described. As described above, the extraction of the noun phrase, the adjective phrase, or the verb phrase here is a premise for calculating the degree of effect, the degree of evaluation aggregation, or the degree of opinion aggregation. Here, an output example of the result of data analysis for a description relating to the current state (items entered in the entry fields 101 and 103 of the entry form 100) is shown.
[0051]
FIG. 6 is a diagram for explaining extraction of noun phrases and adjective phrases in the evaluation analysis. Here, as shown in reverse video, the noun phrase “word registration” is first extracted. Next, as an adjective phrase that modifies the noun phrase “word registration”, an adjective phrase “troublesome” is extracted as shown in a box. In this case, if the frequency of the noun phrase “word registration” is 4, and the level value of the adjective phrase “troublesome” is −1, if there is no other adjective phrase that modifies “word registration”, the degree of effect is − It can be seen that 1 is required.
[0052]
FIG. 7 is a diagram for explaining extraction of adjective phrases and noun phrases in sensitivity analysis. Here, as shown in reverse video, the adjective phrase “correct” is first extracted. Next, as a noun phrase modified by the adjective phrase “correct”, a noun phrase “Japanese” is extracted as shown in a box. In this case, it can be seen that the frequency of the adjective phrase “correct” is 4 and the evaluation intensity is 1.
[0053]
FIG. 8 is a diagram for explaining the extraction of verb phrases and noun phrases in the function request analysis. Here, the verb phrase “setting” is first extracted as shown in reverse video. Next, as a noun phrase modified by the verb phrase “setting”, a noun phrase “dictionary” is extracted as shown in a box. In this case, it is understood that the frequency of the verb phrase “setting” is 4 and the opinion aggregation level is 1.
[0054]
Next, output examples of the results of the evaluation analysis in step S14-1, the sensitivity analysis in step S14-2, and the function request analysis in step S14-3 will be described with specific examples. Here as well, the description regarding the current situation (the items entered in the entry fields 101 and 103 of the entry form 100) will be described.
[0055]
FIG. 9 is a diagram illustrating an output example of the degree of effect of each noun phrase that is the result of the evaluation analysis in step S14-1. In this figure, the vertical axis indicates the degree of noun phrase effect, and the horizontal axis indicates the frequency of the noun phrase. The degree of effect is displayed in the positive area of the vertical axis (upper side of the figure) when the value is positive (when the evaluation is popular), and the vertical axis when the value is negative (when the evaluation is unfavorable) Will be displayed in the negative area (lower side of the figure).
[0056]
Referring to FIG. 9, for example, “conversion accuracy” has a high frequency and a high degree of effect. Therefore, it can be easily recognized visually that the customer is satisfied with the current product. On the other hand, “word registration” shows a negative degree of effect even if the frequency is high, so there is a lot of customer dissatisfaction with the current product, and it should be improved in future product development. Can be easily recognized visually.
[0057]
FIG. 10 is a diagram illustrating an output example of the evaluation intensity for each adjective phrase, which is the result of the sensitivity analysis in step S14-2. In this figure, the vertical axis represents the adjective phrase evaluation intensity, and the horizontal axis represents the frequency of the adjective phrase. Referring to FIG. 10, for example, “difficult” or “correct” has a high frequency and high evaluation aggregation, so that many customers are “difficult” or “correct” with respect to the same point of the same product. You can easily recognize what you feel visually.
[0058]
FIG. 11 is a diagram illustrating an output example of the opinion aggregation level of each verb phrase as a result of the function request analysis in step S14-3. In this figure, the vertical axis indicates the degree of opinion aggregation of the verb phrase, and the horizontal axis indicates the frequency of the verb phrase. Referring to FIG. 11, for example, “installation” has a high frequency and a high degree of opinion aggregation, so it can be visually confirmed that many customers have the same request regarding the same point regarding “installation”. It can be easily recognized.
[0059]
As described above, in the data analysis apparatus 1 according to this embodiment, attention is paid to each noun phrase in the evaluation analysis, and the level of adjective phrases extracted for each noun phrase is aggregated to calculate the degree of effect. . This degree of effect is a numerical value indicating whether the customer's evaluation of the corresponding noun phrase point is favorable or unfavorable. The degree of effect for each noun phrase makes it easy to see how customers are evaluating various aspects of the product, so the results of evaluation analysis can be used to formulate management strategies and marketing strategies. It becomes easy.
[0060]
In the sensitivity analysis, attention is paid to each adjective phrase, and the evaluation intensity is calculated based on the noun phrase modified by each adjective phrase. This evaluation aggregation degree shows numerically whether the way of feeling like the corresponding adjective phrase is concentrated on the same point or distributed over various points. This aggregative evaluation level for each adjective phrase makes it possible to immediately see whether the same feeling is concentrated on the same point, so the results of Kansei analysis can be used to formulate management strategies and marketing strategies. Becomes easier.
[0061]
In function requirement analysis, attention is focused on verb phrases, and the degree of opinion aggregation is calculated based on the noun phrases that each verb phrase modifies. This opinion aggregation level shows numerically whether requests such as corresponding verb phrases are concentrated on the same point or distributed over various points. This level of opinion aggregation for each verb phrase makes it possible to immediately see whether the same request is concentrated on the same point, so the results of functional requirement analysis can be used to formulate management strategies and marketing strategies. Becomes easier.
[0062]
In addition, the degree of effect, evaluation aggregation, and opinion aggregation as a result of the evaluation analysis, sensitivity analysis, and function requirement analysis are output with the vertical axis of the graph and the frequency of the corresponding phrase as the horizontal axis. For this reason, it is possible to easily grasp what many customers feel and demand from visual expression, and use the results of data analysis for the formulation of management strategy and marketing strategy. It becomes easy.
[0063]
Furthermore, the questionnaire which is the object of data analysis in this embodiment is described according to the entry form 100 having entry fields 101-104. Here, descriptions relating to the current state are entered in entry columns 101 and 103, and descriptions relating to future states are entered in entry columns 102 and 104, which are separated from each other. For this reason, it is possible to separate the analysis on the current problem and the analysis of the state desired by the user after solving the problem, so that the results of the data analysis can be used for the formulation of management strategy, marketing strategy, etc. It becomes easy.
[0064]
The present invention is not limited to the above-described embodiment, and various modifications and applications are possible. Hereinafter, modifications of the above-described embodiment applicable to the present invention will be described.
[0065]
In the above embodiment, the customer's questionnaire is provided with the entry form 100 shown in FIG. 2, and the current state and the future state are separated and filled in. However, the current status and the future status are not necessarily clearly separated and the questionnaire is not always filled in. Further, there may be a customer who wants to freely answer a questionnaire without being bound by the entry form 100 of FIG. Unless a filling method such as the filling form 100 in FIG. 2 is used, the current state and the future state are separated in the first place and the questionnaire is not filled.
[0066]
Therefore, data analysis can be performed after the current state and the future state are separated by a method in which any of the following four is arbitrarily combined. In the questionnaire text entered according to the entry form 100 in FIG. 2, it is estimated that the current state and the future state are separated, but the final decision is made after processing by the following method It can be.
[0067]
As a first method, a part where the present tense or past tense is used as a text tense is judged as a description of the current state, and a part where the future tense is used is judged as a description of the future state, Can be separated from the future state. As a second method, since the description about the current state usually appears earlier than the description about the future state, the current state and the future state can be separated according to the description later.
[0068]
As a third method, when focusing on what is described as an inconvenience, the description of the portion with a low degree of effect is grasped as the current state by using the degree of effect obtained by the evaluation analysis in step S14-1. . If there is a portion with a high degree of effect for the same noun phrase in the same sentence, the description of that portion can be grasped as a future state, and the current state and the future state can be separated.
[0069]
As a fourth method, a dictionary as shown in FIG. 12 is prepared as the knowledge dictionary 14b. This dictionary registers words frequently used to describe future states. Among the sentences of the questionnaire, a description of a part including a word / phrase registered in the dictionary shown in FIG. 12 is grasped as a future state. If there is a corresponding description having the same noun phrase in the same sentence, the description of that part can be grasped as the current state, and the current state and the future state can be separated.
[0070]
By clearly distinguishing between the current state and the future state as described above, and analyzing the questionnaire text, it is easy to understand separately the customer satisfaction / dissatisfaction and demand for the products targeted by the questionnaire become able to. As a result, a person who formulates a management strategy and a marketing strategy can easily create an accurate strategy for improving customer satisfaction.
[0071]
In the above embodiment, the text to be analyzed is a questionnaire entered according to the entry form 100 of FIG. 2, and the customer accesses the data analysis apparatus 1 from his / her terminal device 4 and the questionnaire file 14a. Was recorded. On the other hand, the text to be analyzed is not only the data sent by the customer to the data analysis device 1 but also the collected data such as bulletin boards and articles written on the Web server on the Internet 6. There may be. In such a sentence, since the current state and the future state are not separated, application of the above-described separation technique is particularly effective.
[0072]
In the above embodiment, the data analysis device 1 is connected to the terminal device 2 via the LAN 3, and performs data analysis on the questionnaire recorded in the questionnaire file 14a in accordance with an instruction from the terminal device 2, and the analysis result is obtained. It was supposed to be returned to the terminal device 2 of the instruction source. That is, the present invention is realized in a system having a client-server configuration. On the other hand, the present invention may be realized in a stand-alone system assuming that the data analysis apparatus 1 has an input device and a display device.
[0073]
In the above-described embodiment, the program for the data analysis apparatus 1 to perform data analysis on the questionnaire has been described as being stored in the storage device 12 in advance. However, this program may be stored in a computer-readable recording medium such as a CD-ROM or DVD-ROM and distributed independently of the hardware. Alternatively, these processing programs may be stored in a fixed disk device of a Web server device on the Internet and distributed through the Internet.
[0074]
【The invention's effect】
As described above, according to the present invention, it is possible to easily capture subjective information by a text mining technique and perform accurate data analysis. The analysis results also facilitate the formulation of management strategies and marketing strategies.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a system configuration centered on a data analysis apparatus according to an embodiment of the present invention;
FIG. 2 is a diagram showing a questionnaire entry form recorded in a questionnaire file.
FIG. 3 is a diagram illustrating an example of a knowledge dictionary.
FIG. 4 is a flowchart showing processing in the data analysis apparatus according to the embodiment of the present invention.
FIG. 5 is a diagram showing an example of a simple dendrogram display of the subjects classified by the subject analysis.
FIG. 6 is a diagram illustrating extraction of noun phrases and adjective phrases in evaluation analysis.
FIG. 7 is a diagram illustrating extraction of adjective phrases and noun phrases in sensitivity analysis.
FIG. 8 is a diagram for explaining verb phrase and noun phrase extraction in function request analysis;
FIG. 9 is a diagram illustrating an output example of the degree of effect of each noun phrase as a result of the evaluation analysis.
FIG. 10 is a diagram illustrating an output example of evaluation intensity for each adjective phrase, which is a result of sensitivity analysis.
FIG. 11 is a diagram illustrating an output example of opinion aggregation level of each verb phrase as a result of the function request analysis;
FIG. 12 is a diagram showing an example of a dictionary added as a knowledge dictionary in the modification.
[Explanation of symbols]
1 Data analyzer
2 Terminal equipment
3 LAN
4 Terminal equipment
5 Gateway
6 Internet
11 CPU
12 Storage device
13 Communication device
14 File device
14a Questionnaire file
14b Knowledge Dictionary
14c Analysis result database

Claims (10)

Noun phrase dictionary storage means for storing a noun phrase dictionary in which noun phrases that can be included in the sentence to be analyzed are registered;
A modifier phrase storage that stores modifier phrases that are registered by associating modifier phrases that can modify noun phrases registered in the noun phrase dictionary with level values that indicate whether each modifier phrase is semantically preferable or not. Means ,
An analysis object including a description part relating to the current state and a description part relating to a future state related to the present state with reference to the noun phrase dictionary stored in the noun phrase dictionary storage means by instructing the start of sentence analysis Noun phrase extraction means for extracting a noun phrase from each description part included in the sentence to be ,
A modifier phrase extracting means for referring to a modifier phrase dictionary stored in the modifier phrase storage means and extracting a modifier phrase that modifies the noun phrase extracted by the noun phrase extracting means;
Among the sentences to be analyzed, the tense used is the past tense, present tense or future tense, the described position is first or second, or registered in the dictionary stored in the storage means A separation means for separating a description part related to the current state and a description part related to the future state according to whether or not a frequent phrase of the future state is included;
For each descriptive part carved out by the carving means, the level values registered in the modifier phrase dictionary for the modifiers extracted by the modifier extraction means are totaled for each noun phrase extracted by the noun phrase extracting means, An effect level calculating means for calculating the aggregated level value as an effect level for each noun phrase;
A data analysis apparatus comprising: an effect degree output means for outputting the effect degree for each noun phrase calculated by the effect degree calculation means. Noun phrase frequency calculating means for calculating the appearance frequency in each description part included in the sentence to be analyzed of each noun phrase extracted by the noun phrase extracting means for each description part cut by the dividing means ,
2. The effect level output unit outputs the degree of effect calculated by the effect level calculation unit for each noun phrase and the appearance frequency calculated by the noun phrase frequency calculation unit in a visual association. The data analysis device described in 1. Referring to a modifier phrase dictionary stored in the modifier phrase storage means, a modifier phrase extracting means for extracting a modifier phrase that modifies a noun phrase in each description part included in the sentence to be analyzed;
A noun phrase extraction unit that refers to a noun phrase dictionary stored in the noun phrase dictionary storage unit and extracts a noun phrase that is modified by each modifier extracted by the modifier phrase;
Aggregation indicating the degree to which the modifier phrase extracted by the modifier phrase modifier modifies the same noun phrase on the basis of the extracted result of the noun phrase extracted by the noun phrase extractor for each description part segmented by the separator means A degree of aggregation calculating means for calculating the degree for each modifier,
The data analysis apparatus according to claim 1, further comprising an aggregation degree output unit that outputs an aggregation degree for each modifier phrase calculated by the aggregation degree calculation unit. For each description part that has been carved out by the carving means, further comprising a modifier word frequency calculating means for calculating the appearance frequency of each of the description parts included in the sentence to be analyzed of each modifier word extracted by the modifier word extracting means,
The said degree-of-aggregation output means outputs the association degree calculated by said degree-of-aggregation calculation means for each modifier phrase and the appearance frequency calculated by said modifier phrase frequency calculation means in a visually related manner. The data analysis device described in 1. Noun phrase dictionary storage means for storing a noun phrase dictionary in which noun phrases that can be included in the sentence to be analyzed are registered;
A modifier phrase storage means for storing a modifier phrase dictionary in which modifier phrases that can modify noun phrases registered in the noun phrase dictionary are stored ;
An analysis object including a description part relating to the current state and a description part relating to the future state related to the present state with reference to the modifier phrase dictionary stored in the modifier phrase dictionary storage means by instructing the start of sentence analysis A modifier extraction means for extracting a modifier that modifies the noun phrase in each description part included in the sentence ,
A noun phrase extraction unit that refers to a noun phrase dictionary stored in the noun phrase dictionary storage unit and extracts a noun phrase that is modified by each modifier extracted by the modifier phrase;
Among the sentences to be analyzed, the tense used is the past tense, present tense or future tense, the described position is first or second, or registered in the dictionary stored in the storage means A separation means for separating a description part related to the current state and a description part related to the future state according to whether or not a frequent phrase of the future state is included;
Aggregation indicating the degree to which the modifier phrase extracted by the modifier phrase modifier modifies the same noun phrase on the basis of the extracted result of the noun phrase extracted by the noun phrase extractor for each description part segmented by the separator means A degree of aggregation calculating means for calculating the degree for each modifier,
A data analysis apparatus comprising: a degree-of-aggregation output means for outputting the degree of aggregation for each modifier phrase calculated by the degree-of-aggregation calculation means. For each description part that has been carved out by the carving means, further comprising a modifier word frequency calculating means for calculating the appearance frequency of each of the description parts included in the sentence to be analyzed of each modifier word extracted by the modifier word extracting means,
The aggregation level output unit outputs the level of aggregation calculated by the level of aggregation calculation unit for each modifier and the appearance frequency calculated by the modifier frequency calculation unit in a visually correlated manner. The data analysis device described in 1. A noun phrase dictionary that registers noun phrases that can be included in the sentence to be analyzed, and a modifier that can modify the noun phrases registered in the noun phrase dictionary, whether each modifier is semantically preferable or not preferable In a computer apparatus having a dictionary storage means for storing a modifier dictionary registered in association with the level value shown, and having a storage means used as a work area ,
When the start of analysis is instructed, the noun phrase dictionary stored in the dictionary storage means is referred to, and the description part stored in the storage means and the description part relating to the current state and the description part relating to the future state are provided. from each description part included in the text to be analyzed, including extracting noun phrases, stores the noun phrase the extracted in said memory means,
Referring to a modifier dictionary stored in the dictionary storage means, extract a modifier phrase that modifies the noun phrase extracted and stored in the storage means, and stores the extracted modifier phrase in the storage means ,
Among the sentences to be analyzed stored in the storage means, the tense used is the past tense, the present tense or the future tense, the described position is the first or the last, or the dictionary storage means According to whether or not to include a frequent phrase of the future state registered in the stored frequent phrase dictionary, the description part regarding the current state and the description part regarding the future state are separated and stored in the storage means,
Each description part where the carved, aggregated level values registered in the modifier dictionary for each noun phrase is stored in the storage means with said extracted for modifier having stored in the storage means and the extracted The calculated level value is calculated as an effect level for each noun phrase, and the calculated effect level is stored in the storage means.
Data analysis method characterized in that makes output noun phrases for each of Afekuto degree is stored in the storage means and said calculated from the output device of the computer equipment. A dictionary storage means for storing a noun phrase dictionary in which noun phrases that can be included in a sentence to be analyzed are registered, and a modifier phrase dictionary in which modifier phrases that can modify noun phrases registered in the noun phrase dictionary are stored. In addition, in a computer device having storage means used as a work area ,
When the analysis start is instructed, the modifier dictionary stored in the dictionary storage means is referred to, the description part stored in the storage means, and the description part related to the current state and the description part related to the future state in each description part included in the sentence to be analyzed, including extracting the modifier for modifying the noun phrase, it is stored a modifier which the extracted in said memory means,
With reference to the noun phrase dictionary stored in the dictionary storage means, the noun phrases that are modified by each of the modifier phrases that are extracted and stored in the storage means are extracted ,
Among the sentences to be analyzed stored in the storage means, the tense used is the past tense, the present tense or the future tense, the described position is the first or the last, or the dictionary storage means According to whether or not to include a frequent phrase of the future state registered in the stored frequent phrase dictionary, the description part regarding the current state and the description part regarding the future state are separated and stored in the storage means,
Based on the extracted result of the noun phrase, the degree of aggregation indicating the degree of modification of the same noun phrase modified by the extracted extracted and stored in the storage unit for each descriptive part. And the calculated degree of aggregation is stored in the storage means,
Data analysis method characterized in that makes output intensity for each modifier having stored in the storage means and said calculated from the output device of the computer device. A noun phrase dictionary that registers noun phrases that can be included in the sentence to be analyzed, and a modifier that can modify the noun phrases registered in the noun phrase dictionary, whether each modifier is semantically preferable or not preferable A computer device comprising a dictionary storage means for storing a modifier dictionary registered in association with the level value shown,
When an instruction to start analyzing a sentence is given, the noun phrase dictionary stored in the dictionary storage means is referred to, and an analysis target including a description part related to the current state and a description part related to the future state related to the current state is obtained. Noun phrase extraction means for extracting a noun phrase from each description part included in the sentence ,
A modifier phrase extracting unit that extracts a modifier phrase that modifies the noun phrase extracted by the noun phrase extracting unit with reference to the modifier dictionary stored in the dictionary storage unit;
Among the sentences to be analyzed, the tense used is the past tense, present tense or future tense, the described position is first or second, or registered in the dictionary stored in the storage means A separation means for separating a description part about the current state and a description part about the future state according to whether or not a frequent phrase of the future state is included,
For each description part carved out by the carving means, the level values registered in the modifier phrase dictionary for the modifiers extracted by the modifier extraction means are totaled for each noun phrase extracted by the noun phrase extracting means, An effect level calculation means for calculating the total level value as an effect level for each noun phrase, and
A program for functioning as an effect degree output means for outputting an effect degree for each noun phrase calculated by the effect degree calculating means from an output device provided in the computer device . A dictionary storage means for storing a noun phrase dictionary in which noun phrases that can be included in a sentence to be analyzed are registered, and a modifier phrase dictionary in which modifier phrases that can modify noun phrases registered in the noun phrase dictionary are stored. Computer equipment,
Noun phrase dictionary Symbol 憶means for storing the noun phrase dictionary having registered the noun phrase which may be included in the text to be analyzed,
A modifier phrase storage means for storing a modifier phrase dictionary in which modifier phrases that can modify noun phrases registered in the noun phrase dictionary are stored ;
An analysis object including a description part relating to the current state and a description part relating to the future state related to the present state with reference to the modifier phrase dictionary stored in the modifier phrase dictionary storage means by instructing the start of sentence analysis A modifier extraction means for extracting a modifier that modifies the noun phrase in each description part included in the sentence
A noun phrase extraction means for extracting a noun phrase modified by each modifier phrase extracted by the modifier phrase extraction means with reference to the noun phrase dictionary stored in the noun phrase dictionary storage means;
Among the sentences to be analyzed, the tense used is the past tense, present tense or future tense, the described position is first or second, or registered in the dictionary stored in the storage means A separation means for separating a description part about the current state and a description part about the future state according to whether or not a frequent phrase of the future state is included,
Aggregation indicating the degree to which the modifier phrase extracted by the modifier phrase modifier modifies the same noun phrase on the basis of the extracted result of the noun phrase extracted by the noun phrase extractor for each description part segmented by the separator means A degree-of-aggregation calculating means for calculating the degree for each modifier, and
A program for functioning as an aggregation output means for outputting an aggregation degree for each modifier phrase calculated by the aggregation degree calculation means from an output device included in the computer device .

Images