KR20020087612A - Apparatus and method for constructing database of customer relationship management - Google Patents

Abstract

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an apparatus and method for generating a database for customer relationship management, and more particularly, to a relationship between personal information and propensity of customers using information of a customer database that stores and manages personal information and product purchase information. The present invention relates to an apparatus and a method for creating a customer relationship management database (hereinafter referred to as 'CRM-DB'). The present invention analyzes the contents of the customer database to derive the relationship between personal information and propensity of each customer, and automatically generates a high-use knowledge among them, based on the knowledge and externally input knowledge. By building a knowledge base and building a CRM-DB based on that knowledge base, you can operate a CRM-DB more efficiently, which eliminates the need to refer to a large customer database to build a CRM-DB. You can quickly and easily build a CRM-DB and save money and storage space.

Description

Apparatus and method for constructing database of customer relationship management

The present invention relates to an apparatus and method for generating a customer database for customer relationship management, and more particularly, to construct a knowledge base based on the relationship between personal information and propensity of each customer derived from a customer database analysis, An apparatus and method for generating a customer database for customer relationship management, characterized by building a customer relationship management database based thereon.

Customer Relationship Management (CRM) is an information industry term that refers to the methodologies and software that companies need to manage customer relationships in a well-organized way. They usually have Internet service capabilities. In other words, by using the information of the customer database that stores and manages the personal information and goods purchase information of the customer, to grasp the customer relationship in consideration of the relationship between the personal information and the propensity of the customer, it is applied to the corporate marketing or campaign.

This helps the company's marketing department identify their top customers, drive marketing campaigns targeted at them with clear goals, and create quality to drive sales teams. Providing services such as e-mail or product promotion by using such customer relationship management (hereinafter referred to as "CRM") information is called a customer relationship management service (hereinafter referred to as "CRM service").

1 is a schematic block diagram of an apparatus for providing a CRM service according to a conventional embodiment, and FIG. 2 is a flowchart illustrating a process for providing a CRM service according to a conventional embodiment.

Referring to FIG. 1, a conventional CRM service providing apparatus includes a customer DB 10 managing personal information of a customer, purchase information of a customer, a CRM-DB 20 storing and managing customer relationship management information, and a customer DB. (10) analyzing the CRM-DB generation unit 40 to generate the CRM-DB (20), and the DB analysis unit for analyzing the CRM-DB (20) to extract the knowledge of the relationship between the personal information and the propensity of the customer ( 50), a knowledge base 30 that oversees the knowledge, a CRM service provider 60 for providing CRM services to customers with reference to the contents of the knowledge base 30 and the CRM-DB 20, and I Including the control unit 70 for controlling the operation of the CRM-DB generation unit 40, DB analysis unit 50 and CRM service provider 60 by the request of the administrator input through the / F unit 80 It is composed.

The customer DB 10 is a database that manages personal information, product purchase information, etc. of customers, and can be generated by itself or received from an external source. When provided from the outside, the customer DB 10 is provided from various types of customer database sources. It is possible to receive.

Meanwhile, the DB analysis unit 50 uses statistical analysis tools, OLAP (Online Analytic Processing) tools, data mining tools, etc. to extract knowledge, and purchases of customers stored in the CRM-DB 20. The knowledge is extracted by applying techniques such as association rules, classification rules, and prediction models to types.

2 is a flowchart illustrating a process of providing a CRM service according to a conventional embodiment. Referring to Figure 2, the conventional method of providing a CRM service according to this configuration, first gathers the requirements generated from the person who is planning the product marketing and campaign (s10), CRM-DB associated with them from an external customer DB To build (s20). At this time, the required information may include customer's personal specification information, purchase history information, product information, sales region information, and the like. And, by analyzing the built CRM-DB to extract the knowledge based on the purchasing tendency, etc. that appear in common for each customer's personality (s30), to build a knowledge base consisting of these (s40). The CRM service is provided with reference to the knowledge base (S50).

This conventional technique has the following problems.

First, the knowledge base that is finally configured for CRM service is generated by the result of collecting and analyzing the data that is expected to be needed based on the general methodology. Therefore, various types of customer DB are provided from outside. There is a great waste of resources in collecting and analyzing such large databases.

Second, in order to build a CRM-DB with these various types of large-scale DBs, dictionaries such as converting different DBs to a consistent format, correcting missing values, and correcting wrong values through data validation The work requires the generation of refined and consolidated data, which takes a lot of time and effort.

Third, in the conventional case, data mining tools are mainly used to extract knowledge from CRM-DB, and since these data mining tools mainly extract common knowledge level knowledge that we already know, The disadvantage is that knowledge analysts have to spend a lot of time to identify semantic and useful knowledge and register it in the knowledge base.

Accordingly, the present invention has been made to solve the above-mentioned conventional problems, the object of the present invention is to build a knowledge base by analyzing the customer DB centered on the user knowledge, based on the knowledge base customer relationship By automatically establishing a management database (CRM-DB), it is possible to operate CRM-DB more efficiently and to save the cost, storage space and working time for building it. An apparatus and method for generating a database (CRM-DB) are provided.

1 is a schematic block diagram of a CRM service providing apparatus according to a conventional embodiment;

2 is a flowchart illustrating a process of providing a CRM service according to a conventional embodiment;

3 is a schematic block diagram of a CRM service providing apparatus to which a CRM-DB generating apparatus according to an embodiment of the present invention is applied;

4 is a schematic block diagram of a knowledge guide unit according to an embodiment of the present invention;

5 is a flowchart illustrating a process of providing a CRM service after generating a CRM-DB according to an embodiment of the present invention;

5A to 5D are exemplary diagrams of a DB schema for explaining a process of creating a CRM-DB according to an embodiment of the present invention.

6 is a flowchart illustrating a knowledge derivation process according to an embodiment of the present invention;

6A to 6H are exemplary views illustrating a process of deriving knowledge according to an embodiment of the present invention.

7 is a flowchart illustrating a process of deriving knowledge according to another embodiment of the present invention;

7A to 7G are exemplary views illustrating a process of deriving knowledge according to another embodiment of the present invention.

♣ Explanation of symbols for the main parts of the drawing ♣

110: customer DB 120: knowledge base

130: CRM-DB 140: knowledge guide

141: clustering unit 142: display unit

143: statistical significance analysis unit 150: knowledge register

160: knowledge matching unit 170: CRM-DB generation unit

180: CRM service provider 190: control unit

200: I / F part

In order to achieve the above object, the CRM-DB generating apparatus provided in the present invention includes a knowledge base for storing and managing knowledge that is a relationship between personal information and propensity of customers, an interface unit for providing an interface with managers, and an interface unit. Knowledge guide unit for deriving knowledge by analyzing contents of the customer database at the request of a manager; knowledge for registering knowledge directly inputted by the manager through the interface unit and knowledge derived from the knowledge guide unit in the knowledge base. A knowledge matching unit for analyzing the customer database based on a registration unit, knowledge registered in the knowledge base, and extracting a database item corresponding to each knowledge; and a database item for each knowledge extracted from the knowledge matching unit. To build a customer relationship management database with Generating management database (DB-CRM) is characterized by comprising: a.

On the other hand, CRM-DB generation method provided by the present invention to achieve the above object is a first process of deriving the knowledge that is the relationship between personal information and propensity for each customer by analyzing the contents of the customer database by the user knowledge, A second process of constructing a knowledge base based on knowledge derived in step 1 and externally input knowledges, and analyzing the customer database based on knowledge registered in the knowledge base of the second process; And a fourth process of extracting a database item corresponding to knowledge, and a fourth process of constructing a customer relationship management database using the extracted knowledge-based database item.

Hereinafter, with reference to the accompanying drawings will be described in more detail the device and method for generating a CRM-DB of the present invention.

3 is a schematic block diagram of a CRM service providing apparatus to which a CRM-DB generating apparatus according to an embodiment of the present invention is applied.

Referring to FIG. 3, the CRM service providing apparatus to which the CRM-DB generating apparatus of the present invention is applied includes a customer DB 110, a knowledge base 120, a CRM-DB 130, and a knowledge guide unit 140. , The knowledge register 150, the knowledge matcher 160, the CRM-DB generator 170, the CRM service provider 180, the controller 190, and the I / F unit 200. Except for the dual CRM service provider 180 is a configuration of the CRM-DB generating apparatus of the present invention.

The customer DB 110 manages personal information, product purchase information, etc. of the customers, and can be used by itself or by being provided from the outside.

The knowledge base 120 stores and manages knowledge displayed by formalizing the relationship between the customer's disposition inferred by the customer's purchase information of the customer and the personal information of the customer.

The CRM-DB 130 is constructed by data items extracted by applying the knowledge registered in the knowledge base 120 to the customer DB 110 and manages customer information for each knowledge.

The knowledge guide unit 140 derives knowledge by analyzing the contents of the customer DB 110 at the request of the administrator through the I / F unit 200.

The knowledge registration unit 150 registers the knowledge derived from the knowledge guide unit 140 and the knowledge directly input by the administrator through the I / F unit 200 to the knowledge base 120.

At this time, the knowledge registration unit 150 determines the validity of the knowledge by calculating the statistical meaning based on the customer DB 110 in the case of knowledge directly input by the administrator through the I / F unit 200, Only valid knowledge is selected and registered in the knowledge base. That is, by applying the corresponding knowledge (the knowledge directly input by the administrator) to the customer DB 110, if the number of customers satisfying the knowledge is a certain ratio (for example, 70% of the total customer) is recognized as valid knowledge.

In addition, the knowledge registration unit 150 is a new knowledge (k1) to register in the knowledge base 120, the existing knowledge (k2) already registered in the knowledge base and other new knowledge (k3) to register in the knowledge base In the case of having the opposite meaning, only one of the knowledge of the opposite meaning selected by the administrator is registered in the knowledge base. That is, the manager is involved to solve the case where the conflicting meaning between the knowledge to be registered in the knowledge base 120.

Meanwhile, the knowledge registration unit 150 performs a function of automatically combining the newly registered knowledge when there are contents compatible with the previously registered knowledge.

The knowledge matching unit 160 analyzes the customer DB 110 based on the knowledge registered in the knowledge base 120, and extracts a database item corresponding to each knowledge.

The CRM-DB generating unit 170 constructs the CRM-DB 130 with the knowledge-based database items extracted by the knowledge matching unit 160.

Meanwhile, the CRM service provider 180 provides a CRM service to customers using the CRM-DB 130.

4 is a schematic block diagram of the knowledge guide unit 140 according to an embodiment of the present invention. Referring to FIG. 4, the knowledge guide unit 140 of the present invention includes a clustering unit 141, a display unit 142, Statistical meaning analysis unit 143 is included.

The clustering unit 141 clusters the customer DB 110 based on the field information designated by the administrator to derive the relationship between the fields, and then performs a series of processes for displaying the information in a standardized form. In more detail, one field among the fields constituting the customer DB 110 is selected as a field for knowledge derivation, and a predetermined reference field for analyzing the relationship between the field value of the target field and the propensity of the customer is selected. Choose. Then, the relationship between the field value of the target field and the field value of the reference field is derived and the relationship is displayed in a standardized form. In this case, the formalized form generally refers to a conditional sentence syntax for explaining the relationship between the field value of the target field and the field value of the reference field.

The display unit 142 is a device for displaying the clustering result of the clustering unit 141 and may be displayed in a 3D graph or a tree structure.

The statistical meaning analysis unit 143 calculates the statistical meaning of the standardized form in the clustering unit 141, and derives it as knowledge when the value becomes more than a predetermined threshold value. This is similar to the operation of verifying validity of knowledge directly input by the administrator in the knowledge register 150. That is, the clustering unit 141 applies the formalized format to the customer DB 110, and when the number of customers satisfying the knowledge is equal to or greater than a predetermined ratio (for example, 70% of all customers), it is derived as knowledge.

5 is a flowchart illustrating a process of providing a CRM service after generating a CRM-DB according to an embodiment of the present invention. Referring to FIG. 5, a process of generating a CRM-DB using the CRM-DB generating apparatus is as follows.

First, by analyzing the contents of the customer DB based on user knowledge, knowledge that is a relationship between personal information and personality of each customer is automatically derived (s100). At this time, the 'user knowledge' refers to a kind of rule obtained by the experience of the managers in charge of managing the customer, and in order to derive the knowledge for configuring the CRM-DB based on the 'user knowledge' as illustrated in FIGS. 3 and 4. Knowledge guide section 140 is used.

Then, a knowledge base is constructed based on the knowledge and knowledge input from the outside (s200). At this time, in the case of knowledge input from the outside, in order to determine whether the knowledge is valid, a process of calculating the statistical meaning of the knowledge with any one customer DB is performed. This is similar to the process of validating the knowledge directly input by the administrator through the I / F unit 200 in the knowledge register 150 of FIG. 3. Therefore, the specific process is omitted.

On the other hand, when a new knowledge (k1) is to be added to the established knowledge base, the new knowledge (k1) is already registered in the knowledge base (k2) and the other new knowledge to be registered in the knowledge base ( If there is a content that is contrary to k3), the manager who manages the knowledge base is involved to maintain the credibility of the knowledge. At this time, only one of the knowledge of the opposite meaning is selected and registered in the knowledge base.

When the knowledge base is constructed as described above, after analyzing the customer DB based on the knowledge registered in the knowledge base and performing a knowledge matching process for extracting a database item corresponding to each knowledge (s300), the extracted information is then extracted. Build a CRM-DB with knowledge-based database items (s400).

On the other hand, by using the CRM-DB (S500) to provide a specific customer group (S500) such as sending mail (direct mail), customer churn prevention, prospective customer capture, cross-selling (cross-selling).

5A through 5D are exemplary diagrams of a DB schema for explaining a process of creating a CRM-DB according to an embodiment of the present invention. A series of processes illustrated in FIG. 5 will be described with reference to FIGS. 5A through 5D. As follows.

FIG. 5A is an exemplary diagram of a customer management DB, wherein a customer unique number (Cid) / customer name (Name) / address (Address) / phone number (age) / age / job / gender / gender (Sex) field. FIG. 5B is an exemplary diagram of a product management DB. The product unique number (Pid) / producer name (Producer) / product classification (Size) / size (Color) / price (Price) / year of manufacture It includes the (Pyear) / Pmonth / Pday field. FIG. 5C is an exemplary diagram of a sales management DB. The sales number (Sid) / customer unique number (Cid) / product unique number (Pid) / retail-price / date (discount) / discount ( discount / Syear / Smonth / Sday fields.

Meanwhile, FIG. 5D illustrates an example of a CRM-DB generated for sending a DM to a customer group having the highest preference for the record with reference to an external customer DB as shown in FIGS. 5A to 5C for a new record sales purpose.

At this time, in order to generate a CRM-DB as shown in FIG. 5D, first, an external customer DB is analyzed to extract a customer group with high preference for the record, and a feature that is commonly displayed for the customer group with high preference for the record is extracted. do.

For example, if the analysis shows that "People in their 20s or younger in Seoul who bought more than 50,000 won last month and those who bought less than six times a month prefer the record," it is derived from knowledge. To do this, create the following conditions:

<Condition 1>

Customer.age = Customer.age-(Customer.age mod 10)

<Condition 2>

Customer.Last Month Purchase = SUM (retail_price) WHERE smonth = last_month

<Condition 3>

Customer average monthly purchases = AVG (COUNT (*) GROUPBY smonth)

<Condition 1> is a condition to identify the age group who has high preference for record. The preference for record is obtained by subtracting the customer's age (Customer.age) from the customer's age (Customer.age). It is a conditional expression for extracting the age group of customers who have high preference for records by calculating values, such as teenagers, 20s, and 30s.

<Condition 2> is the value of the retail_price field of all records for which the month value of the last month field of the transaction management DB as shown in FIG. Conditional expression to calculate the sum of these values.

<Condition 3> is a conditional expression for calculating an average value of all records having the same Smonth field value of the sales management DB as shown in FIG. 5C, in order to calculate the average number of purchases of each customer.

When the conditional expression necessary to derive the knowledge based on the characteristics of the desired customer groups is generated, knowledge is derived from the combination of the conditional expressions.

In the above example, knowledge of "Customer age group <'20s' and Customer. Last month purchase amount> '50, 000' and Customer.Address =" Seoul "and Customer. Average monthly purchase number <'6 times' → Music preference is derived. . This is generated by the knowledge derivation process s100 of FIG. 5, and may be automatically derived through the knowledge guide unit 140 shown in FIGS. 3 and 4, or manually by an administrator's input. May be derived.

In addition, a process of building a CRM-DB through a knowledge base construction process (s200) for registering such knowledge in a knowledge base and a knowledge matching process (s300) for matching the information registered in the knowledge base with information of each customer. At step S400, the CRM-DB formed by the knowledge condition is as shown in FIG. 5D. That is, referring to Figure 5d, the CRM-DB formed by the knowledge condition includes a customer unique number (Cid) / age group / last month purchase amount / Address (Address) / average monthly purchase count field, each field value is the knowledge condition Extract and store customer information that satisfies all.

FIG. 6 is a flowchart illustrating a process for deriving knowledge according to an embodiment of the present invention, and a process for deriving knowledge about a specific customer group selected by a manager is shown, and FIGS. 6A to 6H illustrate an embodiment of the present invention. It is a specific illustrative view for explaining a process of deriving knowledge according to an example.

Referring to FIGS. 6 and 6A to 6H, a process of deriving knowledge about a specific customer group is as follows.

First, a condition for identifying a specific customer group is selected (s105). At this time, the selected condition is a field name for identifying a specific customer group and a field value of the field. Referring to FIG. 6A, when 'people who buy a cartoon' is selected as a specific customer group for which knowledge is to be derived, the condition for identifying a specific customer group is a customer whose field value of the 'buying magazine' field name is 'cartoon'. .

When the condition is selected as described above, only the customer information satisfying the condition is extracted from the entire customer DB (S110). 6B, in the example of FIG. 6A, only three customers having a field value of 'cartoon' are extracted from the total five customer DBs.

Then, the frequency distribution of the field value of each field of the extracted customer information is checked, and fields having a field value whose number is equal to or greater than a preset threshold value are set as reference fields (S115).

That is, as in the example of FIG. 6C or FIG. 6D, when the number of customers satisfying the condition has a value significantly larger than the number of customers who do not, the field for the condition is set as the reference field. FIG. 6C is a graph showing the result of checking the numbers of field values of the 'income' field, wherein the number of customers whose income is '30' (70) is not the number of customers (3, 15, 15). Has a significantly larger value than. 6D is a graph showing the result of checking the number of field values of the 'car' field indicating whether the vehicle is owned or not, and the number of customers having a car (80) is not the number of customers (12). Has a significantly greater value.

As such, if the random number has a larger value than other values by checking the number of the field values for each field, the field is set as the reference field. 6E is an exemplary diagram showing that the reference field set through such a process is an 'income' field, a 'car' field, and an 'address' field.

On the other hand, by clustering the customer information extracted in the step (s110) based on the reference field, to derive a relationship between the condition selected by the administrator for the specific customer group identification and the field values of the reference field (S120). After displaying the relationship in a formal form (s125), the statistical meaning of the formal form is calculated (s130), and the validity of the formalized form is verified with the result (s135). When only validated form is derived as knowledge (s140).

6F and 6G illustrate examples of displaying clustering results. FIG. 6F illustrates clustering results in a tree structure, and FIG. 6G illustrates clustering results in a three-dimensional graph.

In the case of FIG. 6F, each node represents a statistical meaning of a case in which field values of respective reference fields may be combined, and the form of the combination is “income” of “30” and “car”. ) ", And the reference number" 61 "indicates that 75% of the total target customers satisfy this type of combination, and the reference number" 62 "indicates a customer satisfying this type of combination. For 22% of these total target customers, the reference number '63' indicates that 2% of the total target customers satisfy this form of the combination. Each case represents 1% of the counter customers.

Reference numerals '65 -67 'indicate the statistical meaning of the form of the combination derived from the case of reference numeral' 61 '. For example, reference numeral' 65 'means that' income 'is' 30' , Where 'car' is present, 'and' address' is' Gwanak-gu ', the reference number' 66 'means' income' is' 30 ', and' car ' ) "And" address "is" Gangnam-gu ", and the reference number" 67 "is" income "is" 30 "and" car "is present. And the percentage of customers whose address is Seocho-gu.

At this time, if the ratio is more than a predetermined set value for deriving knowledge, it is derived as knowledge. That is, if the set value is 70%, the form of the combination corresponding to the reference numeral '61' becomes knowledge. If the combination corresponding to the reference number '61' is' income 'is' 30' and 'car' is present ',' income 'is' 30', People with 'car' like 'cartoon' will be derived.

In the case of FIG. 6G, the number of customers satisfying each type of combination is displayed on a three-dimensional graph. In this case, the type of the combination corresponding to the area where the distribution is dense is derived with knowledge. As shown in FIG. 6H, if the type of the combination corresponding to the area where the distribution is dense is "IF (CAR = Y) AND (15,000 <INCOME <16,000) THEN" cartoon "", the result is "IF (CAR = Y) AND ( 15,000 <INCOME <16,000) THEN "Manga Love" knowledge is derived.

FIG. 7 is a flowchart illustrating a process of deriving knowledge according to another embodiment of the present invention. When the administrator does not input information about a specific customer group for deriving knowledge, a process of deriving knowledge from a customer DB is shown. 7A to 7G are exemplary views illustrating a process of deriving knowledge according to the case.

Referring to FIGS. 7 and 7A to 7G, a process of deriving knowledge when a specific customer group is not specified will be described below.

First, any one of the fields constituting the customer DB is selected and set as a target field for deriving knowledge (s155). In the example of FIG. 7A, a field called 'purchase magazine' is set as a target field to derive knowledge.

Then, the customer database is grouped based on the field value of the target field (s160). In FIG. 7B, for this purpose, the customer DB is rearranged by the field value of the field 'purchase magazine'. The group was divided into three groups: 'car', 'purchase magazine' 'music', and 'purchase magazine' 'cartoon'. This is referred to as 'car group', 'music group' and 'cartoon group' for convenience.

When the customers are grouped by the field value of the target field as described above, the field distribution of the field value is checked for each field of the customer information of the group, and the fields having the field value whose number is more than the preset threshold are set as the reference fields. (S165).

That is, FIG. 7C illustrates the distribution of the number of field values of the 'income' field for each group. Referring to FIG. 7C, the number of customers having 'income' of 30 in the 'car group' (70 ) Is significantly higher than the number of customers that do not. On the other hand, in the 'car group', the number of customers with cars (45) is greater than the number of customers (25) without them. Therefore, the 'income' field and the 'car' field may be set as reference fields by this process. FIG. 7D is an exemplary diagram showing that the reference field set through such a process is an 'income' field, a 'car' field, and an 'address' field.

On the other hand, by clustering the customer information for each group based on the reference field to derive the relationship between the field values and the field values of the field selected in the step (s155) (s170), and displays them in a formal form After that, the statistical meaning of the formalized form is calculated (s180). Then, the validity of the formalized form is verified with the result (s185), and only the validated form is derived as knowledge (s190).

7E and 7F are exemplary diagrams for displaying a clustering result, and FIG. 7E is an example of a clustering result in a tree structure, and FIG. 7F is an example of a clustering result in a three-dimensional graph. In this case, the method of analyzing the clustering result shown in FIGS. 7E and 7F is similar to the method of analyzing the clustering result of FIGS. 6F and 6G, and thus a detailed description thereof will be omitted.

On the other hand, FIG. 7G illustrates a case in which the combination of the distribution corresponding to the area where the distribution is dense in the three-dimensional graph shown in FIG. 7F is "IF (CAR = Y) AND (15,000 <INCOME <16,000) THEN" music "". This is an example of the case where knowledge of "IF (CAR = Y) AND (15,000 <INCOME <16,000) THEN" music lover "" is derived.

The above description is merely an example of an embodiment, and the present invention is not limited to the above-described embodiment and can be variously changed within the scope of the appended claims. For example, the shape and structure of each component specifically shown in the embodiment of the present invention can be modified.

The CRM-DB generating apparatus and method thereof of the present invention as described above have the following effects.

First, unlike the conventional method of analyzing a large amount of customer DB to create a CRM-DB and constructing a knowledge base with reference to the CRM-DB, the knowledge and managers automatically derived from the results of arbitrary customer DB analysis By constructing the knowledge base based on the knowledge that is involved and then configuring the CRM-DB based on the knowledge base, there is an effect of reducing the waste of resources required to collect and analyze large databases. In addition, it is not necessary to refer to various types of large-capacity DB in order to configure CRM-DB, thereby saving time and effort for converting these various forms of large-capacity DB into a consistent format. Finally, it analyzes the statistical meaning of each knowledge and registers only the verified knowledge in the knowledge base, thereby reducing the effort of knowledge analysts to register all the knowledge extracted using data mining tools as knowledge for CRM services. Can be.

In other words, it can operate CRM-DB more efficiently and save cost, storage space and working time to build it. Thus, it is suitable to apply to a lightweight customer relationship management system that provides a customer relationship management service even with a small DB.

Claims (9)

An apparatus for creating a customer relationship management database (CRM-DB) in consideration of a relationship between personal information and propensity of customers by using information of a customer database storing and managing customer's personal information and product purchase information, A knowledge base for storing and managing knowledge that is a relationship between personal information and propensity of the customers; An interface unit for providing an interface with the managers, A knowledge guide unit for deriving knowledge by analyzing contents of the customer database at the request of an administrator through the interface unit; A knowledge registration unit for registering knowledge derived from the knowledge guide unit and knowledge directly input by an administrator through the interface unit in the knowledge base; A knowledge matching unit for analyzing the customer database based on knowledge registered in the knowledge base, and extracting a database item corresponding to each knowledge; And a customer relationship management database generation unit for constructing a customer relationship management database using the knowledge-specific database items extracted by the knowledge matching unit. The method of claim 1, wherein the knowledge guide unit A clustering unit configured to derive a relationship between fields by clustering a customer database based on field information designated by the administrator, and displaying the information in a standardized form; A clustering result display unit displaying the clustering result in a 3D graph or a tree structure; A customer comprising a statistical meaning analysis unit which calculates the statistical meaning of the formalized form in the clustering unit with any one customer database and derives it into knowledge when the value is more than a predetermined threshold value Database generation device for relationship management. The method of claim 2, wherein the clustering unit After selecting one field among the fields constituting the customer database as the knowledge derivation target field, selecting a predetermined reference field for analyzing the relationship between the field value of the target field and the propensity of the customer, the field of the target field And a relationship between a value and a field value of the reference field. The method of claim 1, wherein the knowledge registration unit In the case of knowledge directly input by the administrator through the interface unit, After calculating the statistical meaning based on any one customer database to determine the validity of the knowledge, the database generation apparatus for customer relationship management, characterized in that the only valid knowledge is selected and registered in the knowledge base . The method of claim 1, wherein the knowledge registration unit When the new knowledge (k1) to be registered in the knowledge base has a meaning opposite to the existing knowledge (k2) previously registered in the knowledge base and other new knowledge (k3) to be registered in the knowledge base, And registering only one of the knowledge selected by the administrator in the knowledge base. In the method of creating a database (CRM-DB) for creating a customer relationship management database that takes into account the relationship between the personal information and the propensity of the customer by using the information of the customer database that stores and manages the customer's personal information and product purchase information , A first process of analyzing contents of the customer database based on user knowledge and automatically deriving knowledge that is a relationship between personal information and personality of each customer; A second process of constructing a knowledge base based on the knowledge derived from the first process and knowledge input from outside; A third process of analyzing the customer database based on knowledge registered in the knowledge base of the second process and extracting a database item corresponding to each knowledge; And a fourth step of constructing a customer relationship management database using the extracted knowledge-based database items. The method of claim 6, wherein the first process If information about a specific customer group for external knowledge is entered, Step 1-1 to select a condition for identifying the specific customer group, Extracting customer information satisfying the selected condition from the customer database; First to third fields of the field values of field information of the customer information extracted in the first to second stages, and to set fields having field values of which the degree is equal to or greater than a predetermined threshold value as reference fields; Clustering the customer information extracted in step 1-2 based on the reference field to derive a relationship between the field values and the condition selected in step 1-1, and displaying the information in a standardized form. With steps 1-4, And a first to fifth step of calculating a statistical meaning of the standardized form with any one customer database and deriving it into knowledge when the value is above a predetermined threshold value. How to create a database for a project. The method of claim 6, wherein the first process If no information about a specific customer group has been entered for external knowledge, (1-6) selecting any field in the customer database as a target field for deriving knowledge; Grouping the customer database based on field values of the target field; Steps 1-8 of checking field number distribution of field values for each field of customer information for each group, and setting fields having field values whose number is equal to or greater than a predetermined threshold value as reference fields; Clustering customer information for each group based on the reference field to derive the relationship between the field values and the field values of the field selected in the above steps 1-6, and displaying them in a formal format. Steps, Customer relationship management, comprising the steps 1-10 of calculating a statistical meaning of the standardized form with any one customer database and deriving it into knowledge when the value is above a predetermined threshold value; How to create a database for a project. The method of claim 6, wherein the second process is When the new knowledge (k1) to be registered in the knowledge base has a meaning opposite to the existing knowledge (k2) previously registered in the knowledge base and other new knowledge (k3) to be registered in the knowledge base, the knowledge A method of creating a database for customer relationship management, characterized in that the administrator who manages the base participates and selects only one of the knowledge of opposite meanings and registers it in the knowledge base.