JP5329633B2 - Information collection system, information collection method, and program - Google Patents

Description

  The present invention relates to a recommendation technique related to purchase of a product, and more particularly to a technique effective for recommendation information using a social network service.

  In sales sites such as the Internet, search technology that naturally displays recommendations for consumers in searches related to actual purchases is incorporated. For example, a recommendation service is widely known as one of the search techniques.

  The recommendation service is a service that analyzes consumer preferences and displays information that may be of interest to each consumer (recommendation information). Refers to pre-registered consumer preferences, purchase history, etc. Introduce products and services that seem to meet each taste.

  The recommendation information used for the recommendation service is generated based on demographic information including, for example, consumer attribute data (age, gender, income, occupation, etc.), or purchase frequency, purchase history, A technique for generating a purchase amount by performing frequent analysis and priority analysis is known.

  In this kind of recommendation technology, information from an information provider is distributed and collected, and information on a desired subject is collected as needed (for example, refer to Patent Document 1). Personalize search results so that word-of-mouth information is ranked high (for example, refer to Patent Document 2), or use a factor database to analyze and present products that are most suitable for consumers (for example, , See Patent Document 3).

JP 2007-305142 A JP 2009-238007 A JP 2002-108930 A

  However, the present inventors have found that the above-described recommendation information generation technique has the following problems.

  As mentioned above, the recommendation generation technology at sales sites uses demographic information, frequent analysis, and priority analysis, but it does not rely on demographic information for decision making leading to consumer purchases, or There are many cases where there is no correlation with the results of frequent analysis and priority analysis.

  As a result, the accuracy of the recommendation is reduced, the optimal recommendation information cannot be provided to each consumer, and the possibility of quick access to the information desired by the consumer is reduced. It may be difficult to increase the product purchase rate.

  An object of the present invention is to provide a technique capable of performing a highly accurate recommendation without using results such as demographic information, frequent analysis, and priority analysis.

  The above and other objects and novel features of the present invention will be apparent from the description of this specification and the accompanying drawings.

  Of the inventions disclosed in the present application, the outline of typical ones will be briefly described as follows.

  That is, in order to achieve the above-described object, the present invention has a method for ranking preference and purchase awareness in a survey on product preference and purchase awareness, and preference accumulated in social network services regardless of products. A system that provides highly accurate recommendation information without using demographics, information frequent analysis, and priority analysis that can identify individuals by using an autopoietic connection of purchasing awareness.

  The present invention is an information collection system having a server connected via an SNS and a network to which a sales site is connected. In this server, the social graph possessed by the SNS is read out, demographic information is deleted from the social graph, patternized, rankings for preference and purchase intention are output, and output as a conjoint analysis result. Based on, recommend ranking.

  Further, in the present invention, a method using a system for ranking recommendations based on the conjoint analysis result without using demographic, information frequent analysis and priority analysis that can identify an individual, and a computer system function as the system. It can also be applied to programs.

  Among the inventions disclosed in the present application, effects obtained by typical ones will be briefly described as follows.

  Highly recommended recommendation information can be provided.

It is explanatory drawing which shows an example of a structure of the information collection system by one embodiment of this invention. It is a flowchart which shows an example of the purchase intention investigation with respect to a certain goods. It is a flowchart which shows an example of aggregation and analysis of conjoint analysis. It is a flowchart which shows an example of the internal process of a recommendation engine. It is explanatory drawing which shows an example of a conjoint analysis. It is explanatory drawing which shows an example of a social graph. It is explanatory drawing which shows an example of a social graph pattern.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. Note that components having the same function are denoted by the same reference symbols throughout the drawings for describing the embodiment, and the repetitive description thereof will be omitted.

  FIG. 1 is an explanatory diagram showing an example of the configuration of an information collecting system according to an embodiment of the present invention, FIG. 2 is a flowchart showing an example of a purchase intention survey for a certain product, and FIG. 3 is a summary of conjoint analysis. 4 is a flowchart illustrating an example of internal processing of the recommendation engine, FIG. 5 is an explanatory diagram illustrating an example of conjoint analysis, and FIG. 6 is an explanatory diagram illustrating an example of a social graph. FIG. 7 is an explanatory diagram illustrating an example of a social graph pattern.

<Summary of invention>
A first overview of the present invention is an information collection system (information collection system) having a server (recommendation engine 2) connected via a network to which an SNS (SNS6) and a sales site (sales site 8) are connected. System 1). And a server reads the social graph (social graph 7) which SNS has, deletes demographic information from this social graph, patterns it (social graph pattern 142), ranks preference and purchase intention, The conjoint analysis result is output, and the recommendation is ranked based on the conjoint analysis result.

  The second outline of the present invention is information that collects recommendation information of consumers by a computer (information collection system 1) that processes information in response to an inquiry from a sales site (sales site 8) connected to a network. It is a collection method. This information collection method includes a step of reading a social graph (social graph 7) of the SNS (SNS 6), a step of deleting demographic information from the social graph, and a pattern of the social graph (social graph pattern 142). And ranking the intention to purchase, and outputting the result as a conjoint analysis result, and ranking the recommendation based on the conjoint analysis result.

  Furthermore, the third outline of the present invention is a step of reading a social graph (social graph 7) included in an SNS (SNS6) connected to the network, a step of deleting demographic information from the social graph, and a pattern of the social graph. (Social graph pattern 142), ranking preference and purchasing intention, outputting the result as a conjoint analysis result, and ranking the recommendation based on the conjoint analysis result. Let the information collection system 1) execute it.

  Hereinafter, the embodiment will be described in detail based on the above-described outline.

<Configuration example of information collection system>
In the present embodiment, the information collection system 1 includes a recommendation engine 2, a result accumulation storage unit 3, and a conjoint analysis result storage unit 4, as shown in FIG.

  The result accumulation storage unit 3 stores the result indicating the intention in the intention survey 5. The intention survey 5 is a survey that is a basis for conjoint analysis, and is composed of question items that investigate the preference and purchase intention of a product that the SNS (social networking service) 6 user wants to investigate, and the result is a social graph. 7 has SNS6.

  The social graph 7 is so-called human relationship correlation data, and includes, for example, human relationship data representing the connection between people, personal data including personal attributes, and connections between people, things, and contents. The conjoint analysis is an analysis method for measuring individual effects of factors constituting “thing” from the overall evaluation of “thing”.

  The conjoint analysis result storage unit 4 is a collection of information suitable for information retrieval. The recommendation engine 2 receives an inquiry from the sales site 8, and uses the social graph 7 corresponding to the content of the inquiry, and presents recommendation information that matches the closest social graph.

  In addition, various functions for performing the processes shown in FIGS. 2 to 7 below collect information in a program format stored in, for example, a program storage memory (not shown) provided in the information collection system 1. This is realized by execution of a CPU (Central Processing Unit) (not shown) of the system 1.

<Example of purchasing intention survey>
FIG. 2 is a flowchart showing an example of a purchase intention survey for a certain product.

  First, it connects to the SNS 6 via a communication line such as the Internet line (step S101), presents the intention survey 5 that is the basis of the conjoint analysis, and makes an opt-in / opt-out inquiry to the intention survey 5 (step S102). .

  In the process of step S102, if the user of SNS 6 is not permitted, the process ends. If permitted, an inquiry is made as to whether or not the social graph can be provided from the SNS 6 (step S103). If permitted, an answer from the intention survey 5 is accepted (step S104). The answer in the process of step S104 is stored in the result accumulation storage unit 3. If the provision of social graph is not permitted in the process of step S103, the process ends.

<Conjoint analysis and analysis example>
FIG. 3 is a flowchart showing an example of conjoint analysis and analysis.

  First, the opt-in person in the investigation shown in FIG. 2 is identified (step S201), and the result corresponding to the intention is read from the result accumulation storage unit 3 for each opt-in unit (step S202).

  Subsequently, the social graph 7 held by the SNS 6 is read (step S203), and a process identification code is assigned to the information of the social graph 7 (step S204).

  Then, the demographic information is deleted from the social graph 7 to which the process identification code is assigned (step S205), and the social graph 7 is patterned (step S206).

  Subsequently, in the patterned social graph 7, the preference and purchase consciousness are ranked, output as a conjoint analysis result (step S 207), and stored in the conjoint analysis result storage unit 4. Thereafter, the processing in steps S202 to S207 is continued until there is no opt-in person for the intention investigation (step S208).

<Example of internal processing of recommendation engine 2>
FIG. 4 is a flowchart showing an example of internal processing of the recommendation engine 2.

  First, when a recommendation inquiry from the sales site 8 is received (step S301), an inquiry is made as to the opt-in of using the social graph 7 for the recommendation (step S302). When the use of the social graph 7 is permitted, an inquiry about the social graph 7 is made (step S303), and a plurality of conjoint analysis result inquiries that the social graph 7 approximates are made (step S304).

  Then, ranking is performed according to the degree of social graph approximation (step S305), and the recommended ranking is performed according to the rank in the conjoint analysis and the social graph approximation (step S306), and a recommended answer is given (step S306). S307).

<Example of conjoint analysis>
FIG. 5 is an explanatory diagram showing an example of conjoint analysis.

  As shown in the figure, the conjoint analysis result includes a processing identification code 141, a social graph pattern 142, preference information 143, and an analysis social graph 147.

  The preference information 143 includes preference (1) 144, preference (2) 145, and preference (3) 146 as subordinate information, and the social graph for analysis 147 includes person-to-person connection information 148, person-to-thing as subordinate information. It has connection information 149, connection information 14A between people and content, and correlation information 14B between people, things, and content.

<Example of social graph>
FIG. 6 is an explanatory diagram illustrating an example of a social graph.

  The left side of FIG. 6 shows an example of the logical structure of the social graph read from the SNS 6.

  In this case, the social graph includes “person-in-charge data” and “behavior history data”.

  “Human relation data” includes data such as “information for identifying a person”, “demographic information”, and “connection between people”. “Information for identifying a person” is, for example, an ID (identification code) and a password. “Demographic information” is information such as age, gender, income, and occupation, for example. “Person-to-person connection” is information about friendship, community participation, affiliated companies, and the like.

  The “behavior history data” includes “connection between people and things”, “connection between people and content”, “correlation information between people, things, and content” and the like. The “connection between people and goods” is, for example, recommendation information and purchase information, and the “connection between people and content” is information such as browsing information on the Web and action history in the site. “Correlation information of people, things, and contents” is information obtained by analyzing each connection (people, things, contents, etc.), for example.

  Moreover, the right side of FIG. 6 shows an example of a social graph for analysis used by the recommendation engine 2 when collecting information and retrieving information.

  Similarly to the social graph read from the SNS 6, the social graph for analysis is composed of “personal data” and “behavior history data”. “Human relation data” includes the process identification code 141 assigned in the process of step S204 of FIG.

  The person-to-person connection 148 includes information obtained by deleting information for identifying an individual from the “person-to-person connection” information of the “human relation data” in the social graph on the left side of FIG. Also, “information for identifying a person” and “demographic information” in the “personal data” in the social graph on the left side of FIG. 6 are deleted.

  In the “behavior history data” of the social graph on the right side of FIG. 6, the connection 149 between people and things, the connection 14A between people and content, and the correlation information 14B between people, things, and content are “behavior” in the social graph on the left side of FIG. It is the same as “History data”.

<Example of social graph patterning>
FIG. 7 is an explanatory diagram showing an example of the social graph pattern 142.

  FIG. 7 illustrates social link and social link destination attributes. The patterning of the social graph includes social link attributes {direct | indirect} and social link destination attributes {workmate | community | classmate | ...}. The social graph patterning is realized by digitizing these into a bitmap.

  As a result, it is possible to use the survey result of the survey subject who has an idea close to the purchaser, so that highly accurate recommendations can be made.

  Thereby, according to this Embodiment, highly accurate recommendation information can be provided with respect to a consumer by performing recommendation using the social graph which excluded demographic information, and the conjoint analysis result.

  As mentioned above, the invention made by the present inventor has been specifically described based on the embodiment. However, the present invention is not limited to the embodiment, and various modifications can be made without departing from the scope of the invention. Needless to say.

  The present invention is suitable for providing recommendation information on social media.

1 Information Collection System 2 Recommendation Engine 3 Result Accumulation Storage Unit 4 Conjoint Analysis Result Storage Unit 5 Intention Survey 6 SNS
7 Social Graph 8 Sales Site 14 Conjoint Analysis Result 14A Connection Information 14B between People and Content 141B Correlation Information 141 between People, Things, and Content Processing Identification Code 142 Social Graph Pattern 143 Preference Information 144 Preference (1)
145 preferences (2)
146 Preferences (3)
147 Social graph for analysis 148 Connection information between people 149 Connection information between people and things

Claims (15)

An SNS and a server connected via a network to which a sales site is connected;
The server
Read social graph of SNS, delete demographic information from the social graph, pattern it, rank preferences and preferences, output as conjoint analysis results, based on the conjoint analysis results, An information collection system characterized by ranking recommendations. The information collection system according to claim 1,
The server
When reading the social graph, a processing identification code is given to the social graph so that the social graph from which demographic information is deleted can be identified. The information collection system according to claim 1 or 2,
The server
The information is characterized in that ranking is performed according to the degree of approximation of the social graph from the conjoint analysis result, and ranking of recommendations is performed based on the conjoint analysis result and ranking based on the degree of approximation of the social graph. Collection system. In the information collection system according to any one of claims 1 to 3,
The server
An information collecting system characterized in that the social graph is patterned by digitizing a social link attribute and a social link destination attribute into a bitmap. The information collection system according to claim 1,
The server
An information collecting system for reading out the social graph permitted by opt-in. An information collection method for collecting consumer recommendation information by a computer that processes information in response to an inquiry from a sales site connected to a network,
Reading a social graph of the SNS;
Deleting demographic information from the social graph;
Patterning the social graph, ranking preference and purchase intention, and outputting as conjoint analysis results;
And a step of ranking recommendations based on the conjoint analysis result. The information collecting method according to claim 6,
A method for collecting information, comprising: adding a process identification code to the social graph, wherein the social graph from which demographic information has been deleted can be identified. The information collection method according to claim 6 or 7,
The step of ranking the recommendations includes:
From the conjoint analysis result, ranking according to the degree of approximation of the social graph;
A method of ranking recommendations based on the conjoint analysis result and ranking based on the degree of approximation of the social graph. In the information collection method of any one of Claims 6-8,
The social graph patterning
An information collection method characterized by digitizing and mapping a social link attribute and a social link destination attribute. The information collecting method according to claim 6,
The information collecting method, wherein the social graph reads a social graph licensed in opt-in. Reading a social graph of an SNS connected to the network;
Deleting demographic information from the social graph;
Patterning the social graph, ranking preference and purchase intention, and outputting as conjoint analysis results;
A program for causing a computer system to execute a step of ranking a recommendation based on the conjoint analysis result. The program according to claim 11, wherein
The program which makes the said computer system perform the step which assign | provides a process identification code to the said social graph, and makes it possible to identify the said social graph from which the demographic information was deleted. The program according to claim 11 or 12,
The step of ranking the recommendations includes:
From the conjoint analysis result, ranking according to the degree of approximation of the social graph;
And a step of ranking recommendations based on the conjoint analysis result and ranking based on the degree of approximation of the social graph. In the program according to any one of claims 11 to 13,
The social graph patterning
A program characterized in that a social link attribute and a social link destination attribute are digitized and bitmapped. The program according to claim 11, wherein
The social graph reads a social graph licensed in opt-in.

Images