KR20190017597A - Method for analyzing social service data and apparatus using the same - Google Patents

Abstract

According to an embodiment of the present invention, provided is a method for analyzing user response data. The method for analyzing user response data comprises the following steps: receiving at least one text which includes a user response; digitizing the at least one text into a matrix; determining a score of the at least one text using an engine formed to predict the score which represents a positive or negative tendency included in the text based on the digitized matrix of the at least one text; and providing the determined score.

Description

TECHNICAL FIELD [0001] The present invention relates to a method and an apparatus for analyzing user reaction data,

The present invention relates to a method and an apparatus for analyzing user response data, and more particularly, to a method and apparatus for analyzing user response data for calculating an emotion index of text among user response data.

There are a variety of ways in which a company that provides a product or service can determine the response of a customer. In recent years, customer responses can be gathered through various surveys that can be offered online, and product reviews on the providing sites. Marketers can refer to customer responses to products or services collected in a variety of ways to improve on a product or service or to plan marketing to meet their needs.

BACKGROUND OF THE INVENTION [0002] Techniques as a background of the invention have been made in order to facilitate understanding of the present invention. And should not be construed as an admission that the matters described in the technical background of the invention are present in the prior art.

However, since existing customer response collections are collected within the platform provided by the company, there is a possibility that the error of representation may occur due to the bias of the information parameter. Also, as a result of our collection of customer responses, customer responses on the company's customer channel are less than 5% of the total social big data.

As a result, 95 percent of customer responses must be reflected in rational decision-making. To this end, it is necessary to objectify the collective emotion in social media. In other words, it is very important for companies to quantify the opinions of customers about products on the big data of social media with the score of products.

On the other hand, over 70% of respondents in social media are neutral and not negative. Traditional social data analysis services do not take into account the specificity and importance of general consumer responses, not brand guardians or libelous defamers. Therefore, companies can not accurately grasp the needs of the majority of consumers, and can experience difficulties in efficient customer customization.

In other words, with the conventional technology, it is difficult to grasp the 'reason' of the customer purchasing behavior that must be grasped in order to increase the sales of the enterprise. For example, conventional dichotomous classification schemes only divide the customer response into satisfaction or dissatisfaction, making it impossible to reflect refined consumer opinions.

Therefore, the inventors of the present invention have recognized that the meaning of the response of the consumers can be more clearly understood by scoring the degree of the emotion without positively or negatively indicating the texts appearing in the social media.

therefore. SUMMARY OF THE INVENTION It is an object of the present invention to provide a user reaction data analysis method and apparatus capable of more accurately grasping a customer response.

Further, the inventors of the present invention have recognized that in scoring texts appearing in social media, a machine-runable model can be used to score positive or negative emotions as scores. Further, it has been recognized that, in using such a machine-run model, it is more accurate to use the text numerically than to use the text as it is, and has an improved efficiency in operation speed. In addition, when text is used numerically, it is recognized that too many variables can degrade the accuracy or reduce the operation speed.

Another problem to be solved by the present invention is to provide a user-responding system that uses a machine-runable model in which the accuracy and speed of operation are improved in order to score positive or negative emotions in scoring texts displayed in social media And to provide a data analysis method and apparatus.

According to an aspect of the present invention, there is provided a method for analyzing user reaction data according to an embodiment of the present invention. The method includes receiving at least one text that includes a user response, quantifying the at least one text into a matrix, determining a tendency to affirm or deny the text based on the quantized matrix for the at least one text Using an engine configured to predict a score to be represented, determining a score for the at least one text, and providing the determined score.

According to another aspect of the present invention, the step of digitizing at least one text into a matrix comprises the steps of counting occurrence frequencies of words in the text, and processing the frequency of occurrences of words in a matrix to reduce variables while retaining information about the text .

According to still another aspect of the present invention, the step of processing the appearance frequencies of the words into a matrix includes a step of replacing the occurrence frequency of words with a matrix and a step of reducing the dimension of the matrix.

According to still another aspect of the present invention, the step of replacing the occurrence frequencies of words with a matrix may be a step of replacing the occurrence frequencies of words with a matrix using one of tf- idf, word2vec, and doc2vec.

According to another aspect of the present invention, the step of reducing the dimension of the matrix may be a step of reducing the dimension using truncated Singular Value Decomposition (SVD) or Sufficient Reduction (SR).

According to another aspect of the present invention, providing a score comprises providing a score for at least one text in association with information of a text creator.

In accordance with another aspect of the present invention, providing a score in association with information of a text creator comprises providing at least one of the at least one textual content using a textual creator influence computed based on at least one of a follower, a friend, And compensating for the score for the text.

According to another aspect of the present invention, the step of providing a score includes classifying at least one text into at least two identifiers according to a score and visually providing two or more identifiers in a time zone.

In order to solve the above problems, an apparatus for analyzing user reaction data according to an embodiment of the present invention is provided. The apparatus comprising: a communication unit operatively coupled to receive at least one text including a user response; at least one textual unit coupled to the communication unit for obtaining a user response from the communication unit; A processor configured to determine a score for at least one text using an engine configured to predict a score indicative of a tendency for the text to be positive or negative based on a quantized matrix for at least one text and a determined score And an output unit configured to provide the output.

According to another aspect of the present invention, digitizing at least one text into a matrix includes counting occurrences of occurrences of words in the text, replacing occurrences of the occurrences of words with a matrix, and reducing the dimension of the matrix.

The present invention has the effect of providing a user reaction data analysis method and apparatus capable of more accurately grasping a customer response.

Furthermore, the present invention can use a machine-run model in a way that improves accuracy and computation speed, in scoring text appearing in social media, to predict scores of positive or negative emotions.

The effects according to the present invention are not limited by the contents exemplified above, and more various effects are included in the specification.

1 is a schematic diagram for explaining a service using a user response data analysis method according to an embodiment of the present invention.
2 is a schematic block diagram for explaining a user reaction data analysis apparatus according to an embodiment of the present invention.
3 is a schematic flowchart for explaining a method of analyzing user reaction data according to an embodiment of the present invention.
4 is a schematic flowchart for explaining a method of analyzing user reaction data according to another embodiment of the present invention.
5 is a schematic diagram for comparing a user reaction data analysis method and a conventional analysis method according to an embodiment of the present invention.
FIG. 6 is a schematic view for explaining a user interface illustrating an emotion index in the user reaction data analysis method according to an embodiment of the present invention. Referring to FIG.
FIG. 7 is a diagram for comparing the buzz amounts referred to in the social media according to various visualization methods in the user reaction data analysis method according to an embodiment of the present invention.
FIG. 8 is a diagram illustrating a score distribution according to various visualization methods in the method of analyzing user response data according to an exemplary embodiment of the present invention. Referring to FIG.
9 is a diagram illustrating words referred to in a lower score according to various visualization methods in the method of analyzing user reaction data according to an embodiment of the present invention.

BRIEF DESCRIPTION OF THE DRAWINGS The advantages and features of the present invention, and the manner of achieving them, will be apparent from and elucidated with reference to the embodiments described hereinafter in conjunction with the accompanying drawings. It should be understood, however, that the invention is not limited to the disclosed embodiments, but is capable of many different forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, To fully disclose the scope of the invention to those skilled in the art, and the invention is only defined by the scope of the claims.

Although the first, second, etc. are used to describe various components, it goes without saying that these components are not limited by these terms. These terms are used only to distinguish one component from another. Therefore, it is needless to say that the first component mentioned below may be the second component within the technical spirit of the present invention.

Like reference numerals refer to like elements throughout the specification.

It is to be understood that each of the features of the various embodiments of the present invention may be combined or combined with each other partially or entirely and technically various interlocking and driving is possible as will be appreciated by those skilled in the art, It may be possible to cooperate with each other in association.

Various embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

1 is a schematic diagram for explaining a system using a user reaction data analysis apparatus according to an embodiment of the present invention. The system 1000 using the user reaction data analysis apparatus 100 includes a social service utilization device 10, a social service server 20, a user reaction data analysis device 100 and a user reaction data analysis utilization device 30 can do. In this specification, the social service utilizing device 10 and the user reaction data analysis utilizing device 30 are not limited and may be, for example, a general purpose computer, a laptop, a network-connected storage, a streaming server, Tablet devices, smart phones), and the like.

User response data is generated in the social service utilizing device 10. [ Social service data may refer to any textual data created by users of social media or social media such as Facebook, Instagram, blog, web page-based social services, tumblers, and the like. The text data may include text for specific services or goods of the type. The social service server 20 stores and manages the social service data generated in the social service using device 10. [ In FIG. 1, the social service server 20 is shown as a single server for the sake of convenience. However, the present invention is not limited to this, and the social service server 20 may include various implementations such as an operation server and a database. In addition, the user response data may include both text data from a social service, as well as text data including user responses obtainable from the Internet.

The user reaction data analysis apparatus 100 according to an embodiment of the present invention is configured to receive social service data stored in the social service server 20 and analyze social service data or text data. In the present specification, the analysis of the text data is performed by determining an emotion score for a text for a specific service or an item-type merchandise, and for generating the emotion score using the emotion score, text having a certain range of emotion scores, And may include, without limitation, all of a series of tasks that provide visual indicia.

The user reaction data analysis apparatus 100 converts the received text data into numerical values, and preferably performs preprocessing on the digitized text data. Furthermore, the user reaction data analysis apparatus 100 acquires emotion scores for text data using a prediction model configured to predict emotion scores of texts when numerical text data is input. In addition, the user reaction data analysis apparatus 100 may provide the user reaction data analysis corresponding to the request of the user reaction data analysis utilizing device 30 using the obtained emotion score. Specific methods by which the user reaction data analysis apparatus 100 processes text data will be described later with reference to FIGS. 3 and 4. FIG.

The user reaction data analysis utilizing device 30 can request the user reaction data analysis apparatus 100 to analyze the social service data associated with the specific service or the non-goods item. In response to the request, the user reaction data analysis utilizing device 30 can receive and view the analysis result of the social service data including the emotion score from the user reaction data analysis apparatus 100.

2 is a schematic block diagram for explaining a user reaction data analysis apparatus according to an embodiment of the present invention.

The user reaction data analysis apparatus 100 includes a communication unit 110, a user input unit 120, an output unit 130, a memory 140, an interface unit 150, a control unit 160, and a power supply unit 170 can do. The components shown in Fig. 1 are not essential, and an apparatus having more or fewer components may be implemented.

The communication unit 110 may include one or more modules that enable wired / wireless communication between the device and the network in which the device is located. The communication unit 110 transmits / receives signals to / from an external device, for example, a social service server on a communication network such as the Internet. The signal may include various types of data. For example, the communication unit 110 may receive user response data, which is text, from the social service server. The user response data may be user response data associated with a particular service or item.

In various embodiments, the user reaction data analysis apparatus 100 may include a crawler module. The crawler module is a module for collecting user reaction data existing in various social services. The crawler module can filter unnecessary text data such as advertisement / spam. Also, the crawler module can collect user reaction data as well as information about the user who produced the data. Information about the users that are gathered together can be, for example, demographic data. The crawler module can collect text data from different social services. In this case, the crawler module can convert the information about the user and unify it into a certain format. In various embodiments, the crawler module may be implemented separately from the user response data analysis device 100, and the user reaction data analysis device 100 may receive user response data (and the demographic data Can be received.

The user input unit 120 generates input data for controlling the operation of the device. The user input unit 120 may include a key pad dome switch, a touch pad (static / static), a jog wheel, a jog switch, and the like.

The output unit 130 is for generating output related to visual, auditory or tactile sense, and may include a display unit 131, an audio output module 132, and the like. The display unit 131 displays (outputs) information processed in the apparatus. For example, the device displays a user interface (UI) or a graphical user interface (GUI) associated with the system. The display unit 131 may be a liquid crystal display (LCD), a thin film transistor-liquid crystal display (TFT LCD), an organic light-emitting diode (OLED), a flexible display display, and a 3D display. The audio output module 132 may output audio data received from the communication unit 110 or stored in the memory 160. [ The sound output module 132 also outputs sound signals related to functions performed in the apparatus. In various embodiments, the user input 120 or the output 130 may be omitted.

The memory unit 140 may store a program for processing and control of the controller 160 or may perform a function for temporarily storing input / output data. The memory 140 may be a flash memory type, a hard disk type, a multimedia card micro type, a card type memory (e.g., SD or XD memory), a RAM (Random Access Memory), SRAM (Static Random Access Memory), ROM (Read Only Memory), EEPROM (Electrically Erasable Programmable Read-Only Memory), PROM A disk, and / or an optical disk. The device may operate in association with a web storage that performs the storage function of the memory 160 on the Internet.

The interface unit 150 serves as a pathway to all external devices connected to the device. The interface unit 150 receives data from an external device, supplies power to each component in the device, or allows data in the device to be transferred to an external device. For example, a wired / wireless headset port, an external charger port, a wired / wireless data port, a memory card port, a port for connecting a device having an identification module, an audio I / O port, A video input / output (I / O) port, an earphone port, and the like may be included in the interface unit 150.

A controller 160 typically controls the overall operation of the device. For example, processing of data or associated control and processing for displaying processed data. The control unit 160 may include a graphics module 161 for parallel data processing. The graphic module 161 may be implemented in the control unit 160 or separately from the control unit 160.

The power supply unit 170 receives external power and internal power under the control of the controller 160 and supplies power necessary for operation of the respective components.

The various embodiments described herein may be embodied in a recording medium readable by a computer or similar device using, for example, software, hardware, or a combination thereof.

According to a hardware implementation, the embodiments described herein may be implemented as application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays May be implemented using at least one of a processor, controllers, micro-controllers, microprocessors, and other electronic units for performing other functions. In some cases, The embodiments described may be implemented by the control unit 160 itself.

According to a software implementation, embodiments such as the procedures and functions described herein may be implemented with separate software modules. Each of the software modules may perform one or more of the functions and operations described herein. Software code can be implemented in a software application written in a suitable programming language. The software code is stored in the memory 160 and can be executed by the control unit 160. [

The control unit 160 may refine the received text data before analysis. More specifically, the control unit 160 may perform the text normalization and the spam / advertisement filtering process on the received text data as a preprocessing process. Text normalization can include misinterpretation, identification and application of newborn spoken words, elimination of unnecessary characters (special symbols, special characters), and cointegration.

The control unit 160 can calculate the emotion score with respect to the preprocessed text data. In addition, the control unit 160 may use the calculated emotion score to analyze the associated sentences. Opinion leader analysis and trend analysis (traffic, referral). For example, brand / keyword trend analysis can include specific user response traffic analysis, specific brand / keyword referral volume analysis, and so on. The control unit 160 may provide the calculated emotion score and an Opinion Leader score (OLS) in association with each other. The OLS can be calculated based on the number of followers, friends, and subscriptions, which means that the influence of the user who created the text on the social service is quantified in detail.

In various embodiments, the control unit 160 can perform data visualization using the calculated emotion score. Data visualization involves presenting emotional distributions according to demographic characteristics of users. Data visualization also includes opinion leader track, demographic information extraction based on stored data, visualization combined with data analysis results, consumer tracking with strong impact on specific brands across various SNS channels, demographic information (Area / age / sex, etc.).

The control unit 160 can perform data processing and visualization. The processing of the data may include combining the emotion scores obtained for each of the textual data. In various embodiments, the controller 160 may combine emotional scores to derive a distribution of emotional scores, including combinations of emotional score spectra, emotional scores and demographic data, and time series analysis of trends according to emotional scores can do. In various embodiments, an OLS high-opinion leader may indicate the response of other users to the restricted text through an emotional score. This allows us to measure the impact of consumer opinion on a particular service or product rather than the impact of each individual on the macroscopic side (number of followers, page thumbnails, etc.) Can be identified.

In various embodiments, the controller 160 may spectrally display the emotion score obtained from the text to score the consumer's opinion on a particular service or non-product, to provide the text of the highest emotion score.

According to the user reaction data analysis apparatus 100 according to the embodiment of the present invention, when a user such as a marketer and a brand manager receives an evaluation of a product or a service from consumers through the output of the user reaction data analysis apparatus 100 Can be obtained as numerical data from a wide range of sources. Accordingly, the present user reaction data analysis apparatus 100 can provide an insight into the improvement of the current service or product and the development of a new product.

Hereinafter, with reference to FIG. 3, a method of acquiring or calculating emotion score in the method of analyzing user reaction data according to an embodiment of the present invention will be described in detail. 3 is a schematic flowchart for explaining a method of analyzing user reaction data according to an embodiment of the present invention. For convenience of explanation, the components will be described with reference to FIG.

First, the control unit 160 receives at least one text including a user response through the communication unit 110 (S110). The text collected in a social service may be text containing user responses associated with a particular service or product. Further, as described above, the text including the user response may be obtained through a method other than the social service. In addition, the text associated with a service or product may be filtered after the text for a certain period of time has been collected without limitation in the social service. The text may be collected in a plurality of social services, in which case the conversion may be accompanied by a uniform formatting of the text. Also, in various embodiments, the information of the user who created the text including the OLS may be received together.

Next, the controller 160 digitizes at least one text in the form of a matrix (S120). Quantifying text in the form of a matrix is not limited, and may include, for example, creating a Document Term Matrix (DTM). DTM means to quantify the frequency of occurrence of a term in a sentence. Through DTM, multiple texts can be represented by a matrix. When a plurality of texts are input, they have one row for each text, and one column for each term.

The control unit 160 determines a score for the at least one text using an engine configured to predict a score indicative of a tendency toward positive or negative of the text based on the numerical matrix for the at least one text S130).

An engine configured to predict a score (or emotional score) indicative of a positive or negative tendency of the text may be configured to predict the emotional score of the text, for example, via text and a dataset of scores, It can be a learned engine. For example, the engine may be constructed using a known and unrestricted method such as Recurrent Neural Network. Each digitized text data in the engine can be represented by one matrix, and an emotion score having a high probability of having digitized text data through repetitive learning can be predicted. Although not limited, affirmative or negative tendencies of each of the terms of the text data are converted into scores, and the emotion score of the text may be predicted or determined according to the learned probability statistics.

The determined score is provided to the user device (S140). For example, a user device that is provided with a determined score may be provided with a corresponding score of the determined score. In various embodiments, the determined emotion score may be generated and provided with various visualization materials described in conjunction with FIG.

The method of analyzing user response data according to an embodiment of the present invention may differ from the conventional method of analyzing text data in that the text itself is not predicted or determined by a score indicating emotion. The text itself can be learned by a prediction engine similar to an emotion score engine, and the text itself can be input to the prediction engine to determine or predict the emotion score. However, when the text itself is used, the throughput of the data may be relatively large as compared with the numerical text data because of the diversity of the text itself. If the data throughput is relatively large, the amount of training to improve the accuracy of the engine may also increase. Or, if an engine with the same amount of training is used, the prediction accuracy may be lower than using text data that is numerically quantified when the text itself is used as input. As a result, according to the user response data analysis method according to an embodiment of the present invention, the empirical score prediction engine capable of learning can provide more accurate results in the same engine learning amount.

4 is a schematic flowchart for explaining a method of analyzing user reaction data according to another embodiment of the present invention. In FIG. 4, steps S110, S130, and S140 are substantially the same as the steps of the same reference numerals shown in FIG. 3, and redundant explanations are omitted.

Steps S122, S124 and S126 may represent another embodiment of the step of quantifying at least one text described in Fig. 3 in the form of a matrix. Referring to FIG. 4, the appearance frequencies of words in at least one text are counted (S122). For example, step S122 may be to generate a Document Term Matrix (DTM).

For example, if one text T1 is "I am Hong Kil-dong" and the other text T2 is "I am Kim Chul-soo", the matrix generated by the DTM is represented by the following matrix .

I The Hong Gil Dong to be Kim Cheol-Soo T1 One One One One 0 T2 One One 0 One One

However, some words in a word, such as an investigation, may not be as important relative to other words in predicting or calculating emotional scores. That is, the text may include words represented by relatively less important variables. However, if you process them to exclude such words or variables, omissions will occur in the text and these omissions can degrade the accuracy of the emotion score being calculated or predicted. To do this, DTM is converted into a matrix and then reduced by dimensioning, which reduces the throughput of data by reducing the variables used while maintaining accuracy.

To this end, the occurrence frequency of words is replaced with a matrix (S124). The method of replacing the DTM with the matrix is not limited. For example, a method such as tf-idf (Term Frequency - Inverse Document Frequency), word2vec, and doc2vec may be used.

The dimension of the substituted matrix is reduced (S126). For the dimensional reduction of the substituted matrix, a truncated singular value decomposition (SVD) or a sum reduction (SR) scheme may be used. When truncated SVD is used, the following equation 1 can be used.

[Equation 1]

는 truncated SVD를 통해 차원 축소된 X, 즉 SVD변환의 결과물이며,

는 SVD변환을 통해 구해진 오른쪽 eigenvector로 이루어진 매트릭스이며, ∑는 SVD변환을 통해 구해진 eigenvalue로 이루어진 대각행렬이며, V는 SVD변환을 통해 구해진 왼쪽 eigenvector로 이루어진 매트릭스이다. 여기서 차원의 크기 k를 조절함으로써 차원 축소를 통해 잃게 되는 정보량, 감성 스코어 예측과 학습에 용이한 데이터의 크기를 고려할 수 있다.Where X is a matrix generated by DTM, k is the size of the dimension for which X is to be reduced,

Is the result of a dimensionally reduced X, or SVD, transform through a truncated SVD,

Is a matrix of right eigenvectors obtained by SVD transform, Σ is a diagonal matrix of eigenvalues obtained by SVD transform, and V is a matrix of left eigenvectors obtained through SVD transform. By adjusting the size k of the dimension, it is possible to consider the amount of information lost through dimension reduction, the sensitivity score prediction, and the size of data that is easy to learn.

For example, the truncated SVD can be used to substantially reduce the size of the substituted matrix. An engine configured to predict a score (or emotion score) indicative of a tendency toward affirmation or negation of the text based on the reduced-size matrix is used (S130). The use of a diminished matrix can enable faster emotional score prediction and learning than using numerical text data (or a matrix) without diminishing dimensions, and accuracy can be maintained at the same time.

5 is a schematic diagram for comparing a user reaction data analysis method and a conventional analysis method according to an embodiment of the present invention.

Figure 5 (a) schematically shows a conventional dichotomous classification of text. Traditionally, text or consumer responses to specific services or goods collected from social services have been expressed as positive or negative responses. Referring to FIG. 5 (b), the user response data analysis method according to an embodiment of the present invention can spectralize the customer response in comparison with the existing dichotomous classification. Accordingly, it is possible to know how the loyal consumer likes the service or the product through the text having the high emotional score (indicated by the emotion index in Fig. 5), and conversely, You can see if there is any improvement. For example, the text creator 204 may be a customer who created a text with a high emotion score, and the text creator 206 may be a customer who created text with a low emotional score.

Also, in the conventional dichotomous classification, it is difficult to know the elements that are of interest to a group that is not positive or negative. The user reaction data analysis method according to an embodiment of the present invention can grasp what factors are important to a group having a high or low emotional score. For example, the text creators 202 and 208 may be customers who have created a text that is not significantly high or low in emotional score.

FIG. 6 is a schematic view for explaining a user interface illustrating an emotion index in the user reaction data analysis method according to an embodiment of the present invention. Referring to FIG. Figure 6 shows an exemplary user interface that provides a score. In the exemplary user interface, texts of the group having the upper 30% and the group having the lower 30% of the emotion score (indicated by the emotion index in FIG. 6) are displayed. Text 211 is a feedback text for a service or product created by the user. The response indication 212 indicates how many other users were impressed or how much the text 211 was shared with respect to the text 211. The icon 213 indicates the social service in which the text is collected. The positive response indication 214 indicates that the text 211 is classified as a positive response. Emotion score 215 represents the degree of positive reaction determined by the engine. The negative indication (216) indicates that the text is classified as a negative response. Emotion score 217 represents the degree of negative reaction determined by the engine. Emotion scores 215 and 217 may have predetermined maximum and minimum values. Although not limited, emotion scores 215 and 217 may have a value from 0 to 100. [ The text corresponding to the emotion scores 215 and 217 can be reviewed as described above, and the satisfaction or need of the consumer with high positive or negative reaction can be known. Also, in various embodiments, an emotion score for at least one text may be provided in association with the information of the text creator. For example, the information of the text creator may be displayed together in Fig.

The emotional score may also be compensated for using the text creator influence calculated based on at least one of followers, friends, and subscriptions in a social service. For example, social service users who are popular with objective reviews may be weighted. Alternatively, the emotional score may be corrected, as the creator with high influence of the text creator writes the text for the service or the product, and for the text commented on the text, the response is judged to be influenced by the creator of the text of the text .

Hereinafter, various methods provided by refining and visualizing the emotion score in the user reaction data analysis method according to an embodiment of the present invention will be described. The above method can display the customer response on the social service in real time. In the following examples, emotional scores were calculated by selecting services related to entertainment, which is an industry sensitive to consumer reputation for reference.

FIG. 7 is a diagram for comparing the buzz amounts referred to in the social media according to various visualization methods in the user reaction data analysis method according to an embodiment of the present invention. Referring to FIG. 7, the amount of social media buzz (or the number of text frequencies) for the keyword is displayed by using the keywords 3, X, Y, and Z as keywords. Social services that collect text include Twitter, YouTube, Instagram, Facebook, and Naver blogs. In various embodiments, the buzz amount by date for the selected keyword can be queried. Accordingly, the total amount of buzz for the related service or product and the traffic amount of the segmented service can be inquired. Also, the buzz amount for each selected keyword can be compared. Accordingly, the service or the product to be intensively selected can be selected, and the recognition and potential market on the social service can be grasped.

FIG. 8 is a diagram illustrating a score distribution according to various visualization methods in the method of analyzing user response data according to an exemplary embodiment of the present invention. Referring to FIG. The provision of an emotional score in various embodiments may include classifying at least one text into two or more identifiers according to the score. For example, a text having an emotion score of 50 to 100 is classified into an identifier (S1), a text having an emotion score of 0 to 50, and a number of negative 0 to negative 50 is classified into an identifier (S2) The text having the negation of 50 to 100 can be classified into the identifier S3. Of course, the scope for classification can be adjusted.

Referring to FIG. 8, an emotional score distribution for a selected service or product is shown. The identifier S1 may be defined as, for example, a detractor, the identifier S2 as a potential guardian, and the identifier S3 as a protector. Thus, the number of defenders, potential guardians, and current guardians of the service can be grasped. Users of the method of pouring user reaction data according to an embodiment of the present invention can inquire and manage the reputation of a service or a product, and can grasp the crisis quickly by grasping the number of defects.

9 is a diagram illustrating words referred to in a lower score according to various visualization methods in the method of analyzing user reaction data according to an embodiment of the present invention. FIG. 9 is a diagram illustrating an example of contents responded positively according to emotion score, keywords thereof, negative contents and keywords of the user response data analysis method according to an embodiment of the present invention. Referring to FIG. 9, an emotion score is acquired for a text for a specific service or product, a most-referred noun 229 Top 5 is extracted from the group 227 classified by the identifier S 1, and a modifier 230 Can be inquired. Alternatively, the most-referenced noun 231, Top 5, may be extracted from the group 228 classified by the identifier S2, and the modifier 232 may be searched. Accordingly, it is possible to provide an insight on the needs of consumers according to each emotional score interval (Detractor, Potential, Promoter).

For example, Warner One's Potential Protectors are showing the greatest interest in the currently popular member Gang Daniel, and they can express their regrets about the missing Prince of Persia. Identifying the main user reactions to these activities and exposing many of Gang Daniel's "sexy" "amounts" could mean that there is a potential for many potential customers to become Warner One's protectors.

Although the embodiments of the present invention have been described in detail with reference to the accompanying drawings, it is to be understood that the present invention is not limited to those embodiments and various changes and modifications may be made without departing from the scope of the present invention. . Therefore, the embodiments disclosed in the present invention are intended to illustrate rather than limit the scope of the present invention, and the scope of the technical idea of the present invention is not limited by these embodiments. Therefore, it should be understood that the above-described embodiments are illustrative in all aspects and not restrictive. The scope of protection of the present invention should be construed according to the following claims, and all technical ideas within the scope of equivalents should be construed as falling within the scope of the present invention.

Claims (10)

Receiving at least one text including a user response;
Digitizing the at least one text into a matrix;
Determining the score for the at least one text using an engine configured to predict a score indicative of a tendency for the text to be positive or negative based on a numerical matrix for the at least one text; And
And providing the determined score. The method according to claim 1,
Wherein quantifying the at least one text into a matrix comprises:
Counting appearance frequencies of words in the text; And
And processing the frequency of occurrences of the words into a matrix to reduce variables while retaining information about the text. 3. The method of claim 2,
Wherein the step of processing the frequency of appearance of the word into a matrix comprises:
Replacing the occurrence frequency of the word with a matrix; And
And reducing the dimension of the matrix. The method of claim 3,
Wherein replacing the occurrence frequencies of the words with a matrix is a step of replacing the occurrence frequencies of the words with a matrix using one of tf- idf, word2vec, and doc2vec. The method of claim 3,
Wherein the step of reducing the dimension of the matrix is a step of reducing a dimension using truncated singular value decomposition (SVD) or a sum reduction (SR). The method according to claim 1,
Wherein providing the score comprises providing the score for the at least one text in association with the information of the text creator. The method according to claim 6,
The step of providing the score in association with the information of the text creator comprises: calculating a score for the at least one text using the text creator influence calculated based on at least one of follower, friend, and subscription in the social service And compensating the user response data. The method according to claim 1,
Wherein providing the score comprises classifying the at least one text into at least two identifiers according to the score; And
And visually providing the at least two identifiers by time zone. A communication unit configured to receive at least one text including a user response;
A communication unit operatively connected to the communication unit,
Acquiring at least one text including a user response from the communication unit,
Digitizing the at least one text into a matrix,
A processor configured to determine the score for the at least one text, using an engine configured to predict a score indicative of a tendency for the text to be positive or negative based on a numerical matrix for the at least one text; And
And an output configured to provide the determined score. 10. The method of claim 9,
Digitizing the at least one text into a matrix,
Count the frequency of occurrence of words in the text,
Replacing the occurrence frequency of the word with a matrix,
And reducing the dimension of the matrix.

Images