WO2016063403A1 - データ分析システム、データ分析方法、およびデータ分析プログラム - Google Patents
データ分析システム、データ分析方法、およびデータ分析プログラム Download PDFInfo
- Publication number
- WO2016063403A1 WO2016063403A1 PCT/JP2014/078248 JP2014078248W WO2016063403A1 WO 2016063403 A1 WO2016063403 A1 WO 2016063403A1 JP 2014078248 W JP2014078248 W JP 2014078248W WO 2016063403 A1 WO2016063403 A1 WO 2016063403A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- data
- classification
- unclassified
- user
- unit
- Prior art date
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/28—Databases characterised by their database models, e.g. relational or object models
- G06F16/284—Relational databases
- G06F16/285—Clustering or classification
- G06F16/287—Visualization; Browsing
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
Definitions
- the present invention relates to a data analysis system for analyzing data.
- Patent Document 1 discloses a game player matching system that can give a general player with a short game play period an opportunity to play against a specific player.
- Patent Document 2 discloses a matching system that supports selection of a matching range by participating players.
- the amount of content included in the service and the number of users who use the service are usually enormous, and it is difficult for conventional techniques to identify the desired data by processing the enormous data. Therefore, for example, each user has hardly been able to find other users who share a common preference with the user.
- the present invention has been made in view of the above-mentioned problems, and its object is to identify a potential other user who has a high possibility of having the same attribute as the user and to present the data to the user Etc. is to provide.
- a data analysis system includes a classification information reception unit that receives classification information indicating a classification of data from a user via a predetermined input device, and a data group.
- the data classification unit that classifies the classification data and the relationship between the unclassified data included in the data group and the classification information are evaluated based on the classification result by the data classification unit.
- Classification data evaluation unit, trend data selection unit that selects unclassified data according to the classification tendency by the user as trend data from the data group according to the evaluation result by the unclassified data evaluation unit, and others related to the trend data
- a user presenting unit that presents the user to the user via a predetermined output device.
- the data analysis system further includes, for example, an element extraction unit that extracts data elements from classification data based on classification information, and an element evaluation unit that evaluates data elements according to a predetermined criterion.
- the unclassified data evaluation unit can evaluate the relevance by using the data element evaluated by the element evaluation unit as one of the classification results.
- the element evaluation unit for example, sets a predetermined amount of transmission information representing a dependency relationship between the data element and the classification information associated with the classification data including the data element. This data element can be evaluated by using it as one of the criteria.
- the data analysis system may further include, for example, an evaluation storage unit that stores an evaluation result by the element evaluation unit in a predetermined storage device.
- the unclassified data is, for example, data including at least a user's evaluation of an event, and is an emotion of the user who generated the unclassified data, based on the evaluation.
- the apparatus further includes an emotion extraction unit that extracts the emotion for the generated event from the unclassified data, and the trend data selection unit can select the trend data according to the extraction result by the emotion extraction unit.
- the data analysis system further includes, for example, an emotion storage unit that stores data elements included in unclassified data and emotion evaluations for the data elements in association with each other in a predetermined storage device.
- the emotion extraction unit can extract the emotion from the unclassified data by evaluating the unclassified data using the emotion evaluation associated with the data element.
- the data analysis system includes, for example, an invitation information receiving unit that receives, from a user, solicitation information that prompts another user to belong to the community to which the user belongs. And an affiliation information generation unit that generates affiliation information that causes the other user to belong to the community when consent is obtained from the other user.
- the unclassified data evaluation unit calculates a score indicating the strength of the connection between the unclassified data and the classification information based on the classification result. Sex can be evaluated.
- the unclassified data evaluation unit can calculate a score based on a correlation between the first data element and the second data element included in the unclassified data, for example. it can.
- the unclassified data includes at least data related to text
- the unclassified data evaluation unit classifies the association between the sentence included in the text and the classification information. It is possible to evaluate based on the result and evaluate the relevance between the unclassified data and the classified information based on the evaluation result.
- the classification information may be information indicating a classification as to whether or not the user's preference is met, for example.
- the data group includes, for example, a web page
- the data, classified data, and / or unclassified data includes, for example, text, images, audio included in the web page. Or data indicating a moving image or a combination thereof.
- the web page is, for example, a page that provides a social network service
- the data indicating text, an image, a sound, a video, or a combination thereof is, for example, It may be data posted by a user using a social network service.
- a data analysis method includes a classification information reception step of receiving classification information indicating classification of data from a user via a predetermined input device, and a data group By associating the classification information with the classification data, the relationship between the data classification step for classifying the classification data and the unclassified data included in the data group and the classification information is not evaluated based on the classification result in the data classification step.
- a data analysis program includes a classification information reception function that receives classification information indicating classification of data from a user via a predetermined input device, and a data group By associating the classification information with the classification data, the data classification function for classifying the classification data and the relationship between the unclassified data included in the data group and the classification information are not evaluated based on the classification result by the data classification function.
- Classification data evaluation function, trend data selection function that selects unclassified data according to the classification tendency by the user as trend data from the data group according to the evaluation result by the unclassified data evaluation function, and other related to trend data
- a user presenting function for presenting the user to the user via a predetermined output device.
- a data analysis system, a data analysis method, and a data analysis program receive classification information indicating a classification of data from a user, and associate the classification information with classification data included in a data group. And classify the information between the unclassified data included in the data group and the classification information based on the classification result, and select the unclassified data according to the classification tendency by the user according to the evaluation result.
- Other users related to the data can be presented to the user. Therefore, the data analysis system or the like has an effect of identifying a potential other user who has a high possibility of having the same attribute as the user and presenting it to the user.
- FIG. 2 is a schematic diagram illustrating a process performed by the data analysis system 100. As illustrated in FIG. 2, using the example in which each user submits a novel book review (data) to a social network service (hereinafter abbreviated as “SNS”) as a data group, Outline the process.
- SNS social network service
- the user classifies the book information (classification data 2a) that suits his / her preference among the book reviews posted by other users, and classifying information 1a (for example, “good” (Such as pressing the “Like” button), it is possible to classify “book reviews that suit the taste” and “book reviews that do not suit the taste”.
- the data analysis system 100 evaluates the relevance between the other book reviews (unclassified data 2b) that have not yet been given the classification information 1a and the classification information 1a based on the classification result (for example, A score indicating high or low is calculated).
- FIG. 3 is a schematic diagram showing a result of processing executed by the data analysis system 100.
- the data analysis system 100 selects / extracts other book reviews in accordance with the classification tendency by the user from the SNS according to the evaluation result, and posts other selected book reviews. List all users.
- the data analysis system 100 analyzes a large number of book reviews posted to the SNS and captures the meaning expressed in the book reviews, whereby a book review similar to the book review that the user gave the classification information 1a (high score). And other users who have posted similar book reviews can be specified.
- the data analysis system 100 analyzes users and attributes (preference, interest, etc.) by analyzing arbitrary data (text, image, sound, video, etc.) included in a data group (for example, a web page such as SNS). , Values, hobbies, occupations, careers, and the like) can be identified and presented to the user.
- a data group for example, a web page such as SNS.
- FIG. 1 is a block diagram illustrating an example of a main configuration of the data analysis system 100.
- the data analysis system 100 is an information processing system including at least one information processing apparatus (for example, a computer such as a personal computer, a server apparatus, or a mainframe) that can execute a data analysis program including a plurality of processes described below. is there.
- a computer such as a personal computer, a server apparatus, or a mainframe
- the data analysis system 100 is realized by one information processing apparatus (computer)
- a plurality of pieces of information that are arbitrarily distributed and executed in a plurality of processes described below It may be a system including a processing device.
- the data analysis system 100 includes a display (display unit), an input device, a memory, and one or more processors capable of executing one or more programs stored in the memory. It can be particularly preferably realized by a multi-function device (for example, a computer).
- the data analysis system 100 includes a control unit 10 (classification information receiving unit 11, data classification unit 12, element extraction unit 13, element evaluation unit 14, unclassified data evaluation unit 15, evaluation storage unit. 16, the trend data selection part 17, the user presentation part 18, the emotion storage part 19, the emotion extraction part 20, the solicitation information reception part 21, the affiliation information generation part 22), the memory
- the control unit 10 comprehensively controls various functions of the data analysis system 100.
- the control unit 10 includes a classification information reception unit 11, a data classification unit 12, an element extraction unit 13, an element evaluation unit 14, an unclassified data evaluation unit 15, an evaluation storage unit 16, a trend data selection unit 17, a user presentation unit 18, and an emotion A storage unit 19, an emotion extraction unit 20, an invitation information reception unit 21, and an affiliation information generation unit 22 are included.
- the classification information receiving unit 11 receives the classification information 1a indicating the classification of the data 2 from the user via a predetermined input device (for example, the input unit 40). That is, the classification information receiving unit 11 acquires the classification information 1 a from the input unit 40 and outputs the acquired classification information 1 a to the data classification unit 12.
- the classified data 2a and the unclassified data 2b are collectively referred to simply as “data 2”.
- the classification information 1a is information indicating a classification of whether or not the user's preference is met, for example.
- the classification information 1a is “like” for the data 2. It may be information indicating whether or not the intention of “!” (According to user's preference) is shown.
- the classification information 1a is not a binary (binary) flag “whether or not the user's preference is met”, for example, “matched”, “somewhat matched”, “somewhat not matched”, Information (multi-value flag) that classifies the degree of preference in multiple stages, such as “does not match”, may be used.
- the data classification unit 12 classifies the classification data 2a by associating the classification information 1a with the classification data 2a included in the data group.
- the data group may be, for example, a web page providing SNS.
- the classification data 2a may be data indicating, for example, text, an image, a sound, a moving image, or a combination thereof included in the web page.
- the data classification unit 12 outputs the classification result 3a in which the classification data 2a and the classification information 1a are associated with each other to the element extraction unit 13.
- Element extraction unit 13 extracts data element 4a from classification data 2a based on classification information 1a.
- the data element 4a is a keyword (for example, a morpheme) included in the text, a partial image included as a part of an image, a partial sound constituting a part of audio, a frame image constituting a moving image, or the like. It's okay.
- the element extraction unit 13 outputs the data element 4a extracted from the classification data 2a to the element evaluation unit 14.
- the element evaluation unit 14 evaluates the data element 4a according to a predetermined standard. For example, the element evaluation unit 14 uses, as one of the predetermined criteria, a transmission information amount representing a dependency relationship between the data element 4a and the classification information 1a associated with the classification data 2a including the data element 4a. Thus, the data element 4a can be evaluated. For example, when the classification data 2a is text included in a web page, and the element extraction unit 13 extracts a keyword included in the text from the text, the element evaluation unit 14 uses the transmitted information amount to determine the keyword. Each keyword is evaluated by calculating a weight. The element evaluation unit 14 outputs the evaluation result (evaluation result 4b) to the unclassified data evaluation unit 15 and the evaluation storage unit 16.
- the unclassified data evaluation unit 15 evaluates the relationship between the unclassified data 2b included in the data group and the classification information 1a based on the classification result 3a by the data classification unit 12. For example, the unclassified data evaluation unit 15 can evaluate the relevance by using the data element 4a evaluated by the element evaluation unit 14 as one of the classification results 3a.
- the unclassified data evaluation unit 15 is scaled so as to take a value indicating the strength of the association between the unclassified data 2b and the classification information 1a (for example, a value of 0 to 10000). The relationship between the two can be evaluated by calculating (based on the classification result 3a).
- the unclassified data evaluation unit 15 first generates a keyword vector indicating whether or not a predetermined keyword is included in the document.
- the keyword vector is, for example, whether each element of the keyword vector takes a value of “0” or “1”, thereby determining whether or not a predetermined keyword associated with the element is included in the text. This is a vector (bag of words).
- the unclassified data evaluation unit 15 changes the element corresponding to the “price” of the keyword vector from “0” to “1”. Then, the uncategorized data evaluation unit 15 calculates the inner product of the keyword vector (vertical vector) and the weight vector (vertical vector using the weight for each keyword as an element) as shown in the following formula. Score S is calculated.
- s represents a keyword vector
- W represents a weight vector
- T represents transposing a matrix / vector (replaces rows and columns).
- the unclassified data evaluation unit 15 may calculate the score S according to the following formula.
- the unclassified data evaluation unit 15 evaluates the first data element (first keyword) included in the unclassified data 2b (the weight of the first keyword) and the first data element included in the unclassified data 2b. The score may be calculated based on the result of evaluation of the two data elements (second keyword) (the weight of the second keyword) (that is, considering the co-occurrence of keywords). The unclassified data evaluation unit 15 may calculate a sentence score for each sentence included in the text, and may calculate the score based on the sentence score (both will be described later in detail).
- the unclassified data 2b may be, for example, data indicating text, an image, a sound, a moving image, or a combination thereof included in the web page, similar to the classified data 2a.
- the unclassified data evaluation unit 15 outputs the evaluation result (evaluation result 4c) to the trend data selection unit 17.
- the evaluation storage unit 16 stores the evaluation result 4b by the element evaluation unit 14 in a predetermined storage device (for example, the storage unit 30). For example, when the classification data 2a is text included in a web page, and the element extraction unit 13 extracts a keyword included in the text from the text, the evaluation storage unit 16 extracts the keyword extracted by the element extraction unit 13 And the weight of the keyword calculated by the element evaluation unit 14 are stored in the storage unit 30 in association with each other.
- a predetermined storage device for example, the storage unit 30.
- the trend data selection unit 17 selects the unclassified data 2b according to the classification tendency by the user as the trend data 2c from the data group according to the evaluation result 4c by the unclassified data evaluation unit 15. For example, when the uncategorized data 2b is text posted by a user using SNS, and the score is calculated as the evaluation result 4c for each text by the unclassified data evaluation unit 15, the trend data selection unit 17 , (1) a text having a score exceeding a predetermined threshold, or (2) a predetermined number (for example, 100) of text in descending order of score is selected as unclassified data 2b in accordance with the classification tendency by the user, The unclassified data 2b is output to the user presentation unit 18 as trend data 2c.
- the trend data selection unit 17 may select all of the unclassified data 2b as the trend data 2c.
- the user presentation unit 18 presents other users related to the trend data 2 c to the user via the display unit 50.
- the trend data 2c input from the trend data selection unit 17 is text posted by a user who uses SNS
- the user presenting unit 18 lists the users (the other users described above) who posted the text.
- the display information 1b for displaying the other user on the display unit 50 is output to the display unit 50 so as to be possible.
- the emotion storage unit 19 stores the data element 4a included in the unclassified data 2b and the emotion evaluation 4d for the data element 4a in association with each other in a predetermined storage device (for example, the storage unit 30). For example, when the data 2 is text included in a web page, the emotion storage unit 19 searches whether or not a predetermined keyword is included in the text. If included, the emotion storage unit 19 extracts the keyword, and stores the emotion score calculated according to a predetermined criterion in the storage unit 30 in association with the keyword as the emotion evaluation 4d.
- a predetermined storage device for example, the storage unit 30.
- the emotion extraction unit 20 is the emotion of the user who generated the unclassified data 2b.
- the emotion for the event generated based on the evaluation is extracted from the unclassified data 2b.
- the user evaluates “It was interesting” for the event “I read a novel” and based on the evaluation (such as the author's style) has a positive feeling of “I like”
- the user evaluates “It was interesting” for the event “I read a novel” and based on the evaluation (such as the author's style) has a positive feeling of “I like”
- I would recommend it to my family” on a given web page (for example, a page that provides SNS) (See FIGS. 2 and 3).
- the emotion extraction unit 20 determines whether or not the keyword included in the text is stored in the storage unit 30 as the data element 4a.
- a positive value (emotion evaluation 4d) of “+1.2” is associated with the data element 4a of “interesting” and stored in the storage unit 30 in advance by the emotion storage unit 19, the emotion extraction The unit 20 sets “+1.2” as the text extraction result 3b.
- a positive value (emotion evaluation 4d) of “+0.8” is associated with the data element 4a of “let's recommend” (utilized form of “recommend”), and further stored in the storage unit 30 by the emotion storage unit 19.
- the emotion extraction unit 20 outputs the extraction result 3b to the trend data selection unit 17.
- the trend data selection unit 17 determines the trend data according to the evaluation result 4c by the unclassified data evaluation unit 15 and the extraction result 3b. 2c can be selected. For example, the trend data selection unit 17 selects, as the trend data 2c, unclassified data 2b that has a score that exceeds a predetermined threshold and from which positive emotions have been extracted (the extraction result 3b has a positive value). You can do it.
- the solicitation information accepting unit 21 accepts solicitation information 1c that prompts other users to belong to the community to which the user belongs from the user via a predetermined input device (for example, the input unit 40). In other words, the solicitation information reception unit 21 acquires the solicitation information 1 c from the input unit 40 and outputs the acquired solicitation information 1 c to the affiliation information generation unit 22.
- the affiliation information generation unit 22 generates affiliation information 3c that causes the other user to belong to the community when consent is obtained from the other user for affiliation to the community, and the affiliation information 3c is stored in the storage unit 30.
- the community to which the other user belongs is added / changed.
- the input unit (predetermined input device) 40 receives input from the user.
- the input unit 40 may be, for example, a mouse, a keyboard, a touch panel, a voice input microphone, or the like.
- 1 illustrates a configuration in which the data analysis system 100 includes the input unit 40.
- the input unit 40 may be any input device (for example, a portable terminal) connected to the data analysis system 100 so as to be able to communicate with the data analysis system 100. Input interface).
- the display unit (predetermined output device) 50 is a device that displays the processing result by the control unit 10 based on the display information 1b input from the user presentation unit 18.
- the display unit 50 may be a liquid crystal display. 1 illustrates a configuration in which the data analysis system 100 includes the display unit 50.
- the display unit 50 may be any output device (for example, a portable terminal) connected to the data analysis system 100 so as to be communicable. Display).
- the storage unit (predetermined storage device) 30 is a storage device configured by an arbitrary recording medium such as a hard disk, an SSD (silicon state drive), a semiconductor memory, a DVD, and the like, and can control the data analysis system 100.
- An analysis program and arbitrary information used by the data analysis system 100 are stored. 1 illustrates a configuration in which the data analysis system 100 includes the storage unit 30, the storage unit 30 may be any storage device connected to the data analysis system 100 so as to be communicable.
- FIG. 4 is a flowchart illustrating an example of processing executed in the data analysis system 100.
- parenthesized “ ⁇ step” represents each step included in the data analysis method.
- the classification information receiving unit 11 receives the classification information 1a indicating the data classification from the user via a predetermined input device (for example, the input unit 40) (Step 1, hereinafter “Step” is abbreviated as “S”).
- Step 1 hereinafter “Step” is abbreviated as “S”.
- the data classification unit 12 associates the classification information 1a with classification data 2a (for example, text described on the web page) included in the data group (for example, web page), thereby classifying the classification.
- the data 2a is classified (S2, data classification step).
- the element extraction unit 13 extracts the data element 4a from the classification data 2a based on the classification information 1a (S3), and the element evaluation unit 14 uses the data element 4a as a predetermined reference (for example, transmission information).
- the amount is evaluated according to (quantity) (S4).
- the evaluation storage unit 16 stores the evaluation result 4b by the element evaluation unit 14 in a predetermined storage device (for example, the storage unit 30).
- the unclassified data evaluation unit 15 evaluates the relationship between the unclassified data 2b included in the data group and the classification information 1a based on the classification result 3a by the data classification unit 12 (S6, unclassified data evaluation step).
- the trend data selection unit 17 selects the unclassified data 2b according to the classification trend by the user as the trend data 2c from the data group according to the evaluation result 4c by the unclassified data evaluation unit 15 (S7, trend) Data selection step).
- the user presentation unit 18 presents another user related to the trend data 2c to the user via a predetermined output device (for example, the display unit 50) (S8, user presentation step).
- the data analysis method may optionally include not only the above-described processing described with reference to FIG. 4 but also processing executed in each unit included in the control unit 10.
- the unclassified data evaluation unit 15 determines whether the first data element included in the unclassified data 2b is evaluated and the second data element included in the unclassified data 2b is evaluated. Based on this, a score can be calculated. For example, when the first keyword appears in the text, the uncategorized data evaluation unit 15 determines the frequency of appearance of the second keyword in the text (that is, the correlation between the first keyword and the second keyword, or co-occurrence). Taking into account, the score of the text can be calculated.
- the unclassified data evaluation unit 15 uses the correlation matrix (co-occurrence matrix) C representing the correlation (co-occurrence) between the first keyword and the second keyword (not the above [Expression 1]),
- the score S can be calculated according to the formula.
- the correlation matrix C is optimized in advance using a learning data set including a predetermined number of predetermined texts. For example, when a keyword “price” appears in a certain text, a value obtained by normalizing the number of occurrences of other keywords with respect to the keyword between 0 and 1 (that is, a maximum likelihood estimate) is the correlation matrix C. Stored in each element (therefore, the sum for each column of the correlation matrix C is 1).
- the data analysis system 100 can calculate the score in consideration of the correlation between keywords, it is possible to identify other potential users who are likely to share attributes with the user with higher accuracy. Can do.
- the unclassified data evaluation unit 15 can calculate a sentence score for each sentence included in each text, and can calculate the score of the text based on the sentence score.
- the uncategorized data evaluation unit 15 generates a keyword vector indicating whether or not a predetermined keyword is included in the sentence included in the text for each sentence.
- the unclassified data evaluation part 15 calculates a score for every text according to the following formula.
- s s is a keyword vector corresponding to the sth sentence. It should be noted that co-occurrence is taken into account (correlation matrix C is used) in calculating the score according to [Equation 4].
- TFnorm can be calculated as shown in [Formula 5] below.
- TF i represents the appearance frequency (Term Frequency) of the i-th keyword
- s ji represents the j-th element of the i-th keyword vector
- c ji represents the correlation matrix C Of j rows and i columns.
- the unclassified data evaluation unit 15 calculates the above score for each text by calculating the following [Equation 6].
- w i is the i-th element of the weight vector w.
- the data analysis system 100 can calculate a score that correctly reflects the sentence meaning, it is possible to identify other potential users who are likely to share attributes with the user with higher accuracy. it can.
- the data analysis system 100 evaluates each of the data elements 4a included in the unclassified data 2b based on predetermined criteria based on the classification information 1a indicating whether the user's preference is met. To do. Then, the data analysis system 100 calculates a score indicating the strength of the connection between the unclassified data 2b and the classified information 1a based on the evaluation result 4b, and the precision (“matches user's preference). The minimum score that can exceed the target value (target adaptation rate) set with respect to the ratio of the trend data 2c selected as “to the data group” can be specified as the adaptation threshold.
- target adaptation rate target adaptation rate
- the data analysis system 100 sets the adaptation threshold based on the classification information 1a given by the user (the result of human judgment on the past data), and the score exceeding the adaptation threshold. Can be selected as data (trend data 2c) that is likely to match the user's preference, and other users related to the trend data 2c can be presented to the user. .
- the data analysis system 100 can classify the unclassified data 2b by analyzing the current data based on the result of analyzing past data. Thereby, the data analysis system 100 can analyze a user's preference in real time, for example (data to be analyzed does not need to be prepared in advance).
- the data analysis system 100 sorts the scores in descending order. Next, the data analysis system 100 scans the classification information 1a assigned to the classification data 2a in order from the classification data 2a having the maximum score (the rank of the score is first). The ratio of the number of pieces of data to which the classification information 1a is assigned to the number of pieces of data that have been scanned at the present time (matching rate) is sequentially calculated.
- the classification information that “according to taste” is obtained when scanning is completed for data with a score rank of 1st to 20th.
- the data analysis system 100 calculates the matching rate as 0.9 (18/20).
- the number of pieces of data to which the classification information 1a of “according to preference” is given is 35. 100 calculates the precision as 0.875 (35/40).
- the data analysis system 100 calculates all the relevance ratios for the classification data 2a and specifies the minimum score that can exceed the target relevance ratio. Specifically, the data analysis system 100 scans the relevance ratio calculated for the classification data 2a in order from the classification data 2a having the minimum score (score rank is 100th). When the target precision is exceeded, the score corresponding to the precision is specified as the minimum score (fit threshold) that can maintain the target precision.
- the data analysis system 100 determines whether or not the score calculated for the unclassified data 2b that has not been determined whether or not it matches the user's preference exceeds the threshold value for matching, The unclassified data 2b determined as exceeding can be selected as the trend data 2c. Thereby, the data analysis system 100 can analyze a user preference in real time.
- Example applied to data groups other than SNS In order to facilitate understanding, the example in which the data analysis system 100 analyzes data included in the SNS (text posted by other users who use the SNS) has been mainly described.
- the data group other than SNS can be used as a data group, and the data included in the data group can be analyzed.
- the data group may be a group of documents collected in the preparation stage of discovery in a US civil lawsuit.
- the data analysis system 100 is an identifier used by the user (reviewer) for classifying the document (document) included in the document group (sorted document group).
- the classification code (tag) is received as classification information 1a, and the document is classified by associating the classification information 1a with the document (classification data) included in the document group.
- the data analysis system 100 evaluates the relevance between the other documents (unclassified data) included in the document group and the classification information 1a based on the classification result (for example, by calculating a score), and A document conforming to the classification tendency by the reviewer is selected and extracted as trend data 2c according to the evaluation result. Finally, the data analysis system 100 displays a list of persons (other users, for example, related parties (custodians) in the lawsuit) related to the trend data 2c. As a result, the data analysis system 100 can reduce the burden on the reviewer that sorts the documents collected in the discovery preparation stage.
- the data analysis system 100 analyzes text
- the data analysis system 100 can also analyze data other than text.
- the data analysis system 100 analyzes speech
- (1) by recognizing the speech the content of the conversation included in the speech may be converted into characters (text), and the text may be analyzed ( 2)
- the voice data may be analyzed as it is.
- the data analysis system 100 converts speech into text by using an arbitrary speech recognition algorithm (for example, a recognition method using a hidden Markov model), and the processing described above. Similar processing is performed on the text. Thereby, the data analysis system 100 can analyze the voice.
- an arbitrary speech recognition algorithm for example, a recognition method using a hidden Markov model
- the data analysis system 100 extracts partial voices (data elements) included in the voice. For example, when the voice “adjust price” is obtained, the data analysis system 100 extracts partial voices “price” and “adjustment” from the voice, and based on the evaluation result of the partial voice, The relevance between the classified voice (unclassified data 2b) and the classified information 1a can be evaluated. In this case, the data analysis system 100 can separate voices using a time series data classification algorithm (for example, a hidden Markov model, a Kalman filter, a neural network, etc.). Thereby, the data analysis system 100 can analyze the voice.
- a time series data classification algorithm for example, a hidden Markov model, a Kalman filter, a neural network, etc.
- the data analysis system 100 can analyze a video (moving image).
- the data analysis system 100 can identify a person included in the frame image by extracting a frame image included in the video and using an arbitrary face recognition technique.
- the data analysis system 100 uses an arbitrary motion recognition technique (for example, a pattern matching technique may be applied), thereby allowing a partial video included in the video (all frame images included in the video to be displayed).
- the motion (motion) of the person can be extracted from the video including a part of the video.
- the data analysis system 100 can evaluate the relevance between the unclassified video (unclassified data 2b) and the classification information 1a based on the person and / or motion. Thereby, the data analysis system 100 can analyze the video.
- the control block (particularly, the control unit 10) of the data analysis system 100 may be realized by a logic circuit (hardware) formed in an integrated circuit (IC chip) or the like, or using a CPU (Central Processing Unit). It may be realized by software.
- the data analysis system 100 includes a CPU that executes instructions of a data analysis program that is software that implements each function, and a ROM (in which the data analysis program and various data are recorded so as to be readable by a computer (or CPU)).
- a Read Only Memory) or a storage device (these are referred to as “recording media”), a RAM (Random Access Memory) for developing the data analysis program, and the like are provided.
- the object of the present invention is achieved by the computer (or CPU) reading the data analysis program from the recording medium and executing it.
- a “non-temporary tangible medium” such as a tape, a disk, a card, a semiconductor memory, a programmable logic circuit, or the like can be used.
- the data analysis program may be supplied to the computer via any transmission medium (such as a communication network or a broadcast wave) that can transmit the data analysis program.
- the present invention can also be realized in the form of a data signal embedded in a carrier wave in which the data analysis program is embodied by electronic transmission.
- the data analysis program causes a computer to realize a classification information reception function, a data classification function, an unclassified data evaluation function, a trend data selection function, and a user presentation function.
- the classification information reception function, data classification function, unclassified data evaluation function, trend data selection function, and user presentation function are the above-described classification information reception unit 11, data classification unit 12, unclassified data evaluation unit 15, trend data selection It can be realized by the unit 17 and the user presenting unit 18, respectively. Details are as described above.
- the data analysis program uses, for example, a script language such as Python, ActionScript, or JavaScript (registered trademark), an object-oriented programming language such as Objective-C or Java (registered trademark), or a markup language such as HTML5. Can be implemented.
- distributed data analysis including an information processing apparatus including each unit that implements each function implemented by the data analysis program and a server device that includes each unit that implements the remaining functions different from the functions. Systems are also within the scope of the present invention.
- a part or all of a data analysis program capable of providing a function of analyzing data is executed in a server device as the data analysis system 100, and a result of the executed processing is returned to an arbitrary information processing terminal.
- the data analysis system of the present invention can function as a server device that is communicably connected to a user terminal via a network.
- the classification information receiving unit 11 is realized in a user terminal (for example, a smartphone, a personal computer, etc.) provided with a predetermined input device and used by the user, and the classification information 1a received by the computer is the data classification unit 12. , Element extraction unit 13, element evaluation unit 14, unclassified data evaluation unit 15, evaluation storage unit 16, trend data selection unit 17, user presentation unit 18, emotion storage unit 19, emotion extraction unit 20, solicitation information reception unit 21, And transmitted to the server device in which the affiliation information generation unit 22 is realized via the network. Then, the server device receives the classification information 1a, executes the various processes described above, and transmits the execution result (display information 1b) to the user terminal.
- a user terminal for example, a smartphone, a personal computer, etc.
- the server device receives the classification information 1a, executes the various processes described above, and transmits the execution result (display information 1b) to the user terminal.
- the data analysis system of the present invention is realized as a system including the server device and the user terminal.
- the data analysis system associates classification information indicating classification of data from a user via a predetermined input device, and associates the classification information with classification data included in a data group.
- a data classification unit for classifying the classification data an unclassified data evaluation unit for evaluating the relevance between the unclassified data included in the data group and the classification information based on the classification result by the data classification unit,
- the other user related to the unclassified data in accordance with the classification tendency by the user is specified according to the evaluation result by the unclassified data evaluation unit, and the specified other user is specified through the predetermined output device. It can also be expressed as a data analysis system including a user presenting unit that presents to a user.
- the data analysis system includes an extraction unit that extracts a classification document group including a predetermined number of documents from document information as a classification target by the user, and a user for each document included in the classification document group.
- a classification code receiving unit that receives a classification code that is an identifier used when classifying documents
- a database that records keywords selected from the documents included in the classification document group based on the classification code
- It can also be expressed as a data analysis system including a score calculation unit that calculates a score obtained by evaluating the strength of association between a document and a classification code based on a keyword.
- the data analysis system is a data analysis system capable of extracting data related to a predetermined case from a plurality of data acquired from around the vehicle, and whether or not the data analysis system is related to the predetermined case.
- the undecided data and the predetermined data are determined based on the already-determined data for which it is determined whether or not the driver driving the vehicle is related to the predetermined case. It can also be expressed as a data analysis system including a relationship evaluation unit that evaluates a relationship with a case and a data notification unit that notifies undecided data to a driver according to the relationship evaluated by the relationship evaluation unit.
- the present invention can be widely applied to an arbitrary computer such as a personal computer, a server device, a workstation, or a mainframe.
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
Description
図2は、データ分析システム100によって実行される処理の過程を示す模式図である。図2に例示されるように、データ群としてのソーシャルネットワークサービス(Social Network Service、以下「SNS」と略記する)に各ユーザが小説の書評(データ)を投稿している例を用いて、上記処理の過程を概説する。
図1は、データ分析システム100の要部構成の一例を示すブロック図である。データ分析システム100は、以下で説明する複数の処理を含むデータ分析プログラムを実行可能な情報処理装置(例えば、パーソナルコンピュータ、サーバ装置、メインフレームなどのコンピュータ)を、少なくとも1つ含む情報処理システムである。
図4は、データ分析システム100において実行される処理の一例を示すフローチャートである。なお、以下の説明において、カッコ書きの「~ステップ」は、データ分析方法に含まれる各ステップを表す。
前述したように、未分類データ評価部15は、未分類データ2bに含まれる第1データ要素が評価された結果と、当該未分類データ2bに含まれる第2データ要素が評価された結果とに基づいてスコアを算出できる。例えば、未分類データ評価部15は、第1キーワードがテキストに出現した場合、当該テキストにおいて第2キーワードが出現する頻度(すなわち、第1キーワードと第2キーワードとの相関、共起ともいう)を考慮して、当該テキストのスコアを計算できる。
前述したように、未分類データ評価部15は、テキストにそれぞれ含まれるセンテンスごとにセンテンススコアを算出し、当該センテンススコアに基づいて当該テキストのスコアを算出できる。この場合、未分類データ評価部15は、当該テキストに含まれるセンテンスに所定のキーワードが含まれるか否かを示すキーワードベクトルを、当該センテンスごとに生成する。そして、未分類データ評価部15は、下記の式にしたがってスコアをテキストごとに算出する。
前述のように、データ分析システム100は、ユーザの嗜好に合っているか否かの分類を示す分類情報1aに基づいて、未分類データ2bに含まれるデータ要素4aを所定の基準に基づいてそれぞれ評価する。そして、データ分析システム100は、当該評価結果4bに基づいて、上記未分類データ2bと上記分類情報1aとの結びつきの強さを示すスコアを算出し、適合率(「ユーザの嗜好に合っている」として選択された傾向データ2cがデータ群に占める割合)に対して設定された目標値(目標適合率)を超過可能な最小のスコアを、適合しきい値として特定することができる。
説明を理解容易とするために、データ分析システム100がSNSに含まれるデータ(当該SNSを利用する他のユーザが投稿したテキスト)を分析する例を主に説明したが、当該データ分析システム100は、SNS以外をデータ群とし、当該データ群に含まれるデータを分析することもできる。例えば、上記データ群は、米国民事訴訟におけるディスカバリの準備段階において収集されたドキュメント群であってよい。
説明を簡略化するために、データ分析システム100がテキストを分析する例を主に説明したが、当該データ分析システム100は、テキスト以外のデータを分析することもできる。例えば、データ分析システム100が音声を分析する場合、(1)音声を認識することによって当該音声に含まれる会話の内容を文字(テキスト)に変換し、当該テキストを分析してもよいし、(2)音声データをそのまま分析してもよい。
データ分析システム100の制御ブロック(特に、制御部10)は、集積回路(ICチップ)等に形成された論理回路(ハードウェア)によって実現してもよいし、CPU(Central Processing Unit)を用いてソフトウェアによって実現してもよい。後者の場合、データ分析システム100は、各機能を実現するソフトウェアであるデータ分析プログラムの命令を実行するCPU、上記データ分析プログラムおよび各種データがコンピュータ(またはCPU)で読み取り可能に記録されたROM(Read Only Memory)または記憶装置(これらを「記録媒体」と称する)、上記データ分析プログラムを展開するRAM(Random Access Memory)などを備えている。そして、コンピュータ(またはCPU)が上記データ分析プログラムを上記記録媒体から読み取って実行することにより、本発明の目的が達成される。上記記録媒体としては、「一時的でない有形の媒体」、例えば、テープ、ディスク、カード、半導体メモリ、プログラマブルな論理回路などを用いることができる。また、上記データ分析プログラムは、当該データ分析プログラムを伝送可能な任意の伝送媒体(通信ネットワークや放送波等)を介して上記コンピュータに供給されてもよい。本発明は、上記データ分析プログラムが電子的な伝送によって具現化された、搬送波に埋め込まれたデータ信号の形態でも実現され得る。
データを分析する機能を提供可能なデータ分析プログラムの一部または全部が、データ分析システム100としてのサーバ装置において実行され、当該実行された処理の結果が任意の情報処理端末に返される構成であってもよい。すなわち、本発明のデータ分析システムは、ユーザ端末とネットワークを介して通信可能に接続されたサーバ装置として機能することができる。
本発明は上述したそれぞれの実施の形態に限定されるものではなく、請求項に示した範囲で種々の変更が可能であり、異なる実施の形態にそれぞれ開示された技術的手段を適宜組み合わせて得られる実施の形態についても、本発明の技術的範囲に含まれる。さらに、各実施の形態にそれぞれ開示された技術的手段を組み合わせることにより、新しい技術的特徴を形成できる。
Claims (15)
- データの分類を示す分類情報を、所定の入力装置を介してユーザから受け付ける分類情報受付部と、
データ群に含まれる分類データに前記分類情報を対応付けることによって、当該分類データを分類するデータ分類部と、
前記データ群に含まれる未分類データと前記分類情報との関連性を、前記データ分類部による分類結果に基づいて評価する未分類データ評価部と、
前記ユーザによる分類傾向に則した未分類データを、前記未分類データ評価部による評価結果に応じて、前記データ群から傾向データとして選択する傾向データ選択部と、
前記傾向データに関連する他のユーザを、所定の出力装置を介して前記ユーザに提示するユーザ提示部とを備えたデータ分析システム。 - 前記分類情報に基づいて前記分類データからデータ要素を抽出する要素抽出部と、
前記データ要素を所定の基準にしたがって評価する要素評価部とをさらに備え、
前記未分類データ評価部は、前記要素評価部によって評価された前記データ要素を前記分類結果の1つとして用いることによって、前記関連性を評価することを特徴とする請求項1に記載のデータ分析システム。 - 前記要素評価部は、前記データ要素と当該データ要素を含む分類データに対応付けられた分類情報との依存関係を表わす伝達情報量を、前記所定の基準の1つとして用いることによって、当該データ要素を評価することを特徴とする請求項2に記載のデータ分析システム。
- 前記要素評価部による評価結果を所定の記憶装置に格納する評価格納部をさらに備えたことを特徴とする請求項2または3に記載のデータ分析システム。
- 前記未分類データは、事象に対するユーザの評価を少なくとも含むデータであり、
前記未分類データを生成したユーザの感情であって、前記評価に基づいて生じた前記事象に対する感情を、当該未分類データから抽出する感情抽出部をさらに備え、
前記傾向データ選択部は、前記感情抽出部による抽出結果にさらに応じて、前記傾向データを選択することを特徴とする請求項1から4のいずれか一項に記載のデータ分析システム。 - 前記未分類データに含まれるデータ要素と当該データ要素に対する感情評価とを対応付けて、所定の記憶装置に格納する感情格納部をさらに備え、
前記感情抽出部は、前記データ要素に対応付けられた前記感情評価を用いて前記未分類データを評価することによって、前記感情を当該未分類データから抽出することを特徴とする請求項5に記載のデータ分析システム。 - 前記ユーザが所属するコミュニティに所属するように前記他のユーザを促す勧誘情報を、前記所定の入力装置を介して当該ユーザから受け付ける勧誘情報受付部と、
前記所属について前記他のユーザから承諾を得られた場合、当該他のユーザを前記コミュニティに所属させる所属情報を生成する所属情報生成部とをさらに備えたことを特徴とする請求項1から6のいずれか一項に記載のデータ分析システム。 - 前記未分類データ評価部は、前記未分類データと前記分類情報との結びつきの強さを示すスコアを前記分類結果に基づいて算出することによって、前記関係性を評価することを特徴とする請求項1から7のいずれか一項に記載のデータ分析システム。
- 前記未分類データ評価部は、前記未分類データに含まれる第1データ要素と第2データ要素との相関に基づいて前記スコアを算出することを特徴とする請求項8に記載のデータ分析システム。
- 前記未分類データは、テキストに関するデータを少なくとも含み、
前記未分類データ評価部は、前記テキストに含まれるセンテンスと前記分類情報との関連性を、前記分類結果に基づいて評価し、当該評価結果に基づいて、前記未分類データと当該分類情報との関連性を評価することを特徴とする請求項1から9のいずれか一項に記載のデータ分析システム。 - 前記分類情報は、前記ユーザの嗜好に合っているか否かの分類を示す情報であることを特徴とする請求項1から10のいずれか一項に記載のデータ分析システム。
- 前記データ群は、ウェブページを含み、
前記データ、分類データ、および/または未分類データは、前記ウェブページに含まれるテキスト、画像、音声、もしくは動画、またはこれらの組み合わせを示すデータを含むことを特徴とする請求項1から11のいずれか一項に記載のデータ分析システム。 - 前記ウェブページは、ソーシャルネットワークサービスを提供するページであり、
前記テキスト、画像、音声、もしくは動画、またはこれらの組み合わせを示すデータは、前記ソーシャルネットワークサービスを利用するユーザによって投稿されたデータであることを特徴とする請求項12に記載のデータ分析システム。 - データの分類を示す分類情報を、所定の入力装置を介してユーザから受け付ける分類情報受付ステップと、
データ群に含まれる分類データに前記分類情報を対応付けることによって、当該分類データを分類するデータ分類ステップと、
前記データ群に含まれる未分類データと前記分類情報との関連性を、前記データ分類ステップにおける分類結果に基づいて評価する未分類データ評価ステップと、
前記ユーザによる分類傾向に則した未分類データを、前記未分類データ評価ステップにおける評価結果に応じて、前記データ群から傾向データとして選択する傾向データ選択ステップと、
前記傾向データに関連する他のユーザを、所定の出力装置を介してユーザに提示するユーザ提示ステップとを含む、データ分析システムにおいて実行されるデータ分析方法。 - データの分類を示す分類情報を、所定の入力装置を介してユーザから受け付ける分類情報受付機能と、
データ群に含まれる分類データに前記分類情報を対応付けることによって、当該分類データを分類するデータ分類機能と、
前記データ群に含まれる未分類データと前記分類情報との関連性を、前記データ分類機能による分類結果に基づいて評価する未分類データ評価機能と、
前記ユーザによる分類傾向に則した未分類データを、前記未分類データ評価機能による評価結果に応じて、前記データ群から傾向データとして選択する傾向データ選択機能と、
前記傾向データに関連する他のユーザを、所定の出力装置を介してユーザに提示するユーザ提示機能とを、コンピュータに実現させるデータ分析プログラム。
Priority Applications (3)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US15/521,184 US20170351747A1 (en) | 2014-10-23 | 2014-10-23 | Data analysis system, data analysis method, and data analysis program |
PCT/JP2014/078248 WO2016063403A1 (ja) | 2014-10-23 | 2014-10-23 | データ分析システム、データ分析方法、およびデータ分析プログラム |
JP2016542301A JP6043460B2 (ja) | 2014-10-23 | 2014-10-23 | データ分析システム、データ分析方法、およびデータ分析プログラム |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
PCT/JP2014/078248 WO2016063403A1 (ja) | 2014-10-23 | 2014-10-23 | データ分析システム、データ分析方法、およびデータ分析プログラム |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2016063403A1 true WO2016063403A1 (ja) | 2016-04-28 |
Family
ID=55760470
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2014/078248 WO2016063403A1 (ja) | 2014-10-23 | 2014-10-23 | データ分析システム、データ分析方法、およびデータ分析プログラム |
Country Status (3)
Country | Link |
---|---|
US (1) | US20170351747A1 (ja) |
JP (1) | JP6043460B2 (ja) |
WO (1) | WO2016063403A1 (ja) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116304277A (zh) * | 2023-03-01 | 2023-06-23 | 深圳一资源网络平台有限公司 | 一种基于ai的智能匹配方法、系统和存储介质 |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP7049010B1 (ja) * | 2021-03-02 | 2022-04-06 | 株式会社インタラクティブソリューションズ | プレゼンテーション評価システム |
CN118364115B (zh) * | 2024-06-20 | 2024-09-06 | 潍坊学院 | 一种产品设计信息分类系统 |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2005181928A (ja) * | 2003-12-24 | 2005-07-07 | Fuji Xerox Co Ltd | 機械学習システム及び機械学習方法、並びにコンピュータ・プログラム |
JP2011170786A (ja) * | 2010-02-22 | 2011-09-01 | Nomura Research Institute Ltd | 文書分類システムおよび文書分類プログラムならびに文書分類方法 |
JP2012079247A (ja) * | 2010-10-06 | 2012-04-19 | Hironori Abe | 口コミ情報管理システム及び口コミ情報管理プログラム |
JP2013182338A (ja) * | 2012-02-29 | 2013-09-12 | Ubic:Kk | 文書分別システム及び文書分別方法並びに文書分別プログラム |
Family Cites Families (9)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8620723B1 (en) * | 2011-07-20 | 2013-12-31 | Ourgroup, Inc. | System and method for generating a community contribution index |
JP5738785B2 (ja) * | 2012-02-21 | 2015-06-24 | Kddi株式会社 | ユーザ嗜好に合った批評者を推薦するレコメンドプログラム、装置及び方法 |
JP2013235507A (ja) * | 2012-05-10 | 2013-11-21 | Mynd Inc | 情報処理方法、装置、コンピュータプログラムならびに記録媒体 |
US9165328B2 (en) * | 2012-08-17 | 2015-10-20 | International Business Machines Corporation | System, method and computer program product for classification of social streams |
US20140052540A1 (en) * | 2012-08-20 | 2014-02-20 | Giridhar Rajaram | Providing content using inferred topics extracted from communications in a social networking system |
JP6028582B2 (ja) * | 2013-01-16 | 2016-11-16 | 大日本印刷株式会社 | サーバ装置、プログラム及び通信システム |
US9788777B1 (en) * | 2013-08-12 | 2017-10-17 | The Neilsen Company (US), LLC | Methods and apparatus to identify a mood of media |
US9773269B1 (en) * | 2013-09-19 | 2017-09-26 | Amazon Technologies, Inc. | Image-selection item classification |
US9705801B2 (en) * | 2013-10-07 | 2017-07-11 | Facebook, Inc. | Systems and methods for mapping and routing based on clustering |
-
2014
- 2014-10-23 US US15/521,184 patent/US20170351747A1/en not_active Abandoned
- 2014-10-23 WO PCT/JP2014/078248 patent/WO2016063403A1/ja active Application Filing
- 2014-10-23 JP JP2016542301A patent/JP6043460B2/ja not_active Expired - Fee Related
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2005181928A (ja) * | 2003-12-24 | 2005-07-07 | Fuji Xerox Co Ltd | 機械学習システム及び機械学習方法、並びにコンピュータ・プログラム |
JP2011170786A (ja) * | 2010-02-22 | 2011-09-01 | Nomura Research Institute Ltd | 文書分類システムおよび文書分類プログラムならびに文書分類方法 |
JP2012079247A (ja) * | 2010-10-06 | 2012-04-19 | Hironori Abe | 口コミ情報管理システム及び口コミ情報管理プログラム |
JP2013182338A (ja) * | 2012-02-29 | 2013-09-12 | Ubic:Kk | 文書分別システム及び文書分別方法並びに文書分別プログラム |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116304277A (zh) * | 2023-03-01 | 2023-06-23 | 深圳一资源网络平台有限公司 | 一种基于ai的智能匹配方法、系统和存储介质 |
CN116304277B (zh) * | 2023-03-01 | 2023-12-15 | 张素愿 | 一种基于ai的智能匹配方法、系统和存储介质 |
Also Published As
Publication number | Publication date |
---|---|
US20170351747A1 (en) | 2017-12-07 |
JP6043460B2 (ja) | 2016-12-14 |
JPWO2016063403A1 (ja) | 2017-04-27 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
Cappallo et al. | New modality: Emoji challenges in prediction, anticipation, and retrieval | |
Das et al. | Applications of artificial intelligence in machine learning: review and prospect | |
US9563652B2 (en) | Data analysis system, data analysis method, data analysis program, and storage medium | |
Myers et al. | Racial population projections and reactions to alternative news accounts of growing diversity | |
CN108628833B (zh) | 原创内容摘要确定方法及装置,原创内容推荐方法及装置 | |
Jerez-Fernandez et al. | Show me the numbers: Precision as a cue to others’ confidence | |
JP5359399B2 (ja) | テキスト分析装置および方法、並びにプログラム | |
Sanchez-Cortes et al. | In the mood for vlog: Multimodal inference in conversational social video | |
Zangerle et al. | Leveraging affective hashtags for ranking music recommendations | |
JP2009099088A (ja) | Snsユーザプロファイル摘出装置、摘出方法並びに摘出プログラム、及び該ユーザプロファイルを利用する装置 | |
JP2014149713A (ja) | 画像評価装置 | |
JP7280705B2 (ja) | 機械学習装置、プログラム及び機械学習方法 | |
Abouelenien et al. | Gender-based multimodal deception detection | |
US9058328B2 (en) | Search device, search method, search program, and computer-readable memory medium for recording search program | |
JP5905652B1 (ja) | データ評価システム、データ評価方法、およびデータ評価プログラム | |
JP6043460B2 (ja) | データ分析システム、データ分析方法、およびデータ分析プログラム | |
Chen et al. | Semantic Space models for classification of consumer webpages on metadata attributes | |
Yousefi et al. | Examining Multimodel Emotion Assessment and Resonance with Audience on YouTube | |
JP5942052B1 (ja) | データ分析システム、データ分析方法、およびデータ分析プログラム | |
Mamo | Big data and innovative research methods | |
Marshall | A latent allocation model for brand awareness and mindset metrics | |
JP2016177690A (ja) | サービス推薦装置およびサービス推薦方法並びにサービス推薦プログラム | |
Pérez-Rosas et al. | Gender differences in deceivers writing style | |
CN109977231B (zh) | 一种基于情感衰变因子的抑郁情绪分析方法 | |
Chaurasia et al. | Predicting mental health of scholars using contextual word embedding |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
ENP | Entry into the national phase |
Ref document number: 2016542301 Country of ref document: JP Kind code of ref document: A |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 14904265 Country of ref document: EP Kind code of ref document: A1 |
|
WWE | Wipo information: entry into national phase |
Ref document number: 15521184 Country of ref document: US |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 14904265 Country of ref document: EP Kind code of ref document: A1 |