JP2001282985A - Sales information analyzing method - Google Patents

Abstract

PROBLEM TO BE SOLVED: To grasp purchase behavior by combining the feature of a merchandise and the feature of a customer purchasing the merchandise by combining sales information including purchase class information and merchandise information and analyzing a sales situation. SOLUTION: An analyzed category, an analysis key and a purchase class type are designated. They are stored in a memory and required information is extracted from sales information and merchandise information in accordance with a designated condition. The analysis key and the purchase class type are made to be a reference and sales results are summed up. It is detected whether the significant feature can be recognized in the constitution rate of the summed-up result of the sales results.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to merchandise analysis, and more particularly to a method of analyzing sales information while combining an analysis key for arbitrary merchandise information and purchase layer information.

[0002]

2. Description of the Related Art A method disclosed in Japanese Patent Application Laid-Open No. 11-7587 is a method for retrieving and classifying information on the sales performance of a product based on the product information of the product. According to this method, sales results can be searched, classified, and totaled using the product information.

[0003]

According to the above-mentioned prior art, sales results are searched, classified, and totaled based on product information, which is a specific feature of registered products. However, in the method of analyzing based on the product information, it is not possible to analyze what kind of featured product the customer bought.

An object of the present invention is to provide a method for analyzing what kind of merchandise is to be purchased for each purchasing layer by using the product information and the purchasing layer information mutually.

[0005]

The present invention provides means for providing sales information, means for providing product information, and storage means for managing the provided sales information and the product information.
Means for analyzing managed information.

Further, the present invention is characterized in that it has a step of detecting that a characteristic is recognized in the composition ratio of the sales performance in an arbitrary analysis key and an arbitrary purchase layer.

Further, the present invention is characterized in that it has a step capable of detecting that the composition ratio of the sales record changes with a characteristic in a time series in an arbitrary analysis key and an arbitrary purchase layer.

[0008]

DESCRIPTION OF THE PREFERRED EMBODIMENTS One embodiment of the present invention will be described below with reference to the drawings.

FIG. 1 is a configuration diagram of a network for collecting and analyzing sales information according to the present embodiment. The sales information providing client 1, the product information providing client 2, and the information analysis client 4 are connected to the information management server 3 via the network 5. The sales information providing client 1 provides sales information including purchase layer information, and the information management server 3
To be stored. The product information providing client 2 provides product information and stores it in the information management server 3. The information analysis client 4 extracts necessary information from the information management server 3, analyzes the information by a predetermined method, and obtains an analysis result.

FIG. 2 shows the storage device 10 of the information management server 3.
FIG. The storage device 10 stores a product information DB 12, a sales information DB 14, and a category DB 16. The product information DB 12 stores information on product attributes.
The sales information DB 14 stores sales information including purchase layer information. The category DB 16 stores products belonging to the category.

FIG. 3 is a PAD diagram showing a procedure of a process for registering product information. The product information provider uses the product information providing client 2 to specify the product code of the product to be registered, and sets its analysis key and scale (step 1).
00). Next, the designated product code and the set analysis key and scale of the product code are stored in the product information DB 12 on the information management server 3 (step 102).

FIG. 4 is a diagram showing a data structure of the product information DB 12. As shown in FIG. The product information DB 12 indicates an analysis key and a scale for each product, and the registered product code 104
, An analysis key 106 corresponding to the product code, and a scale 108 corresponding to the analysis key.

FIG. 5 is a PAD showing a procedure of a process for registering sales information. The sales information provider uses the sales information providing client 1 and obtains, for each purchase unit 115, the purchased product code 116, the purchased quantity of the product code 116, The sales information is obtained so that the purchase layer information can be understood (step 110). Next, the edited sales information is stored in the sales information DB 14 on the information management server 3 (step 11).
2).

FIG. 6 is a diagram showing a data structure of the sales information DB 14. As shown in FIG. The sales information DB 14 indicates, for each purchase unit 115, the purchased product code 116, the quantity 117, the purchase layer type 118 of the customer who purchased at that time, and the attributes thereof. Here, when the purchase unit 115 is 1, the product codes 116 of the purchased product are 1 and 2, and the product with the product code 116 is purchased in a quantity of 3, and the purchase layer type of the customer who purchased the product is At 118, the attribute relating to the purchasing tier A is 1, the attribute relating to the purchasing tier B is 1, and the attribute relating to the purchasing tier C is 1.

FIG. 7 is a PAD diagram showing a procedure for registering category information. The information analyst uses the information analysis client 4 to set a category 131 and a product code 132 belonging to the category (step 12).
0). Next, the set category 131 and product code 132 are stored in the category DB on the information management server 3.
16 (step 122).

FIG. 8 shows the data structure of the category DB 16. The category DB 16 has a table for each analyst 130, and each table has a product code 132 of a product belonging to the category 131 for each category 131.
Is shown. Here, the product codes 132 of the products belonging to category 1 are 1, 2, 3, 4, and 5.

FIG. 9 shows an example of using product information and sales information.
PA showing a procedure of a process for determining whether a significant characteristic is recognized in the sales situation by a purchase layer in an arbitrary analysis key
FIG. First, the information analyst uses the information analysis client 4 to set a category 131 to be analyzed. When it is necessary to specify the analysis key 106 to be detected, the analysis key 106 is also specified. At this time, the analysis key 106 corresponding to the product belonging to the designated category 131 is extracted and the information analysis client 4
And may be used as determination information when specifying. Also,
When it is necessary to specify the purchase layer type 118 to be detected, the purchase layer type 118 is also specified. Analysis key 13
1 and the purchase tier type 118 are not specified, it is assumed that all the analysis keys 106 and all the purchase tier types 118 corresponding to the products belonging to the specified category 131 are specified. Category 1 specified
31, the analysis key 106 and the purchase layer type 118 are stored in a memory (not shown) on the information management server 3 (step 150). Next, the analysis key 106 and the purchase layer type 118 stored in the memory on the information management server 3
Steps 154 to 158 are repeated while some of the combinations are not analyzed (Step 1
52). First, from the sales results of each product belonging to the specified category 131 on the memory on the information management server 3, as for the purchase layer information, the specified purchase layer type 118 is extracted from the sales information DB 14 for the specified purchase layer type 118. Store it in memory on the management server. The analysis key 1 of each product code 132 on the memory on the information management server 3
The scale 108 relating to 06 is extracted from the product information DB 12 and stored in the memory on the information management server 3. Each product code 132 of each sales record on the memory on the information management server 3 and the analysis key 106 of the product code 132
Is stored in the memory of the information management server 3 after being associated with the scale 108 (step 154). next,
The sales results, which are stored in the memory on the information management server 3 and are associated with the scale 108 of each product, are stored for each attribute of the specified purchase layer type 118 and for each scale 108 of the specified analysis key 106. Then, the sales performance quantity 178 is cross tabulated and stored in the memory on the information management server 3 (step 156). Next, the sales performance for each attribute of the purchasing tier type 118 and the scale 108 of the analysis key 106 stored in the memory on the information management server 3 are converted into the specified purchasing tier type 118 and the specified analysis. Whether or not there is a relationship between the key 106 and the sales trend is verified by the chi-square test, the test result is stored in the memory on the information management server 3, and displayed on the information analysis client 4 (step 158). For example, Chi-square test is an introduction to data analysis (Keio University Press)
It is described in. The i-th cell of the analysis key scale 196 and the attribute 198 of the purchase layer type in the cross tabulation table of sales results for each attribute 198 of the purchase layer type and each analysis key scale 196
Of the j-th cell, and the sum of each row and each column is n _i ., N. _J , respectively, and the total sum is n. In addition, the number of cells in the row of the cross tabulation table is k, and the number of cells in the column is m. Expected value of each square is calculated by _{E ij = n i .n. J} / n. The ratio of the difference between the actual value and the expected value in each cell is (O _ij −E
_ij ) ² / E Calculate by _ij and add the sum of all cells

[0018]

(Equation 1)

Calculated as follows. Also, the degree of freedom (k−1) × (m
-1) is calculated by Using a chi-square distribution table, and any significant level, the degree of freedom to determine the x ^2. x ² <x ² ₀
In this case, the test result indicates that there is a relationship between the purchase layer type 118 and the analysis key 131.

FIG. 10 is a diagram showing the structure of data stored in the memory in step 154. The sales results of the products belonging to the designated category 170 are displayed. The scale 176 for each product code 174 of the specified analysis key 172 is extracted. In addition, the attribute of the specified purchase layer type is displayed. Here, category 17
0 is 1 and the product code 174 is 1, 2 and 5
Is displayed. Analysis key 1
When 72 is 2, the scale of each product is 1, 10,
It is 7. Further, when the purchase layer type is the purchase layer A180, the customer attributes are 1, 1, and 4, respectively.

FIG. 11 is a diagram showing the structure of data stored in the memory in step 156. Specified category 190, analysis key 192, purchase layer type 194
Each table has a table, and the result of cross tabulation of the sales volume for each scale 196 and each attribute 198 of the purchase layer type is displayed. Here, the scale 196 is the scale 1, and the attribute 198 of the purchase layer type is 3 and the sales performance quantity 178 of the purchase layer 1 is 3.

FIG. 12 is an example of a screen displayed on the information analysis client 4 for designating a category, an analysis key, and a purchase layer type. The category 190 to be analyzed is designated by the category 200. The analysis key 202 is used to specify an analysis key 192 used for analysis. Purchasing layer type 2
04 designates a purchase layer type 194 to be used for analysis. The analysis key 192 used for the analysis is determined by the scale 206.
Is specified. The attribute 208 specifies the attribute 198 of the purchase layer type 194 used for analysis. By specifying the candidate selection 210, options related to the specified candidate selection are displayed in the candidates 212.

In FIG. 9, it may be determined whether a significant feature is recognized in the sales situation in the scale 108 of the specified specific analysis key 106 or the specified attribute of the specific purchase layer type 118. First, a specific analysis key 106
When the scale 108 is specified, in addition to the analysis key 106,
The scale 108 of the analysis key 106 is designated. When the attribute of the specific purchase layer type 118 is specified, the attribute of the purchase layer type 118 is specified (step 150). Next, the analysis key 106 stored in the memory on the information management server 3
Steps 154 to 158 are repeated while there is a combination that has not been analyzed among the combinations of the purchase tier types 118 (step 152). First, from the sales results of each product belonging to the category 131 specified in the memory on the information management server 3, with regard to the purchase layer information,
The sales information DB 1 for the specified purchase stratum type 118
4 and stored in the memory on the information management server. The scale 108 relating to the analysis key 106 of each product code 132 on the memory on the information management server 3 is extracted from the product information DB 12 and stored in the memory on the information management server 3. Each product code 132 of each sales record on the memory on the information management server 3 and the product code 132
Is stored in the memory of the information management server 3 after being associated with the scale 108 of the analysis key 106 (step 154). Next, from the sales results stored in the memory on the information management server 3 and associated with the scale 108 of each product, the specified purchase layer type 118 is selected.
The sales performance quantity 178 is cross-tabulated for each attribute of and for the scale 108 of the specified analysis key 106. further,
If the attribute of the purchase layer type 118 is specified, the attributes of the specified purchase layer type 118 and other attributes are totaled. If the scale 108 of the analysis key 106 is specified, the scale 108 of the specified analysis key 106 and the other scales 108 are totaled and stored in the memory on the information management server 3 (step 156). Next, each attribute of the aggregated purchase layer type 118 stored in the memory on the information management server 3 and the scale 10 of the analysis key 106
The sales results for each of the 8 are tested by a chi-square test to determine whether there is a relationship between the specified purchase layer type 118 and the specified analysis key 106, and the test results are stored in a memory on the information management server 3. Stored in the information analysis client 4
(Step 158).

In FIG. 9, it may be determined whether or not a feature of the sales status or a feature of an arbitrary product is significant in a time series. that time,
Category 131 to be analyzed, analysis key 1 if necessary
06. In addition to specifying the purchase layer type 118, the information analysis client 4 specifies the attribute of the purchase layer type 118 or the scale 108 of the analysis key 106 to be used in a time-series analysis. It is stored in the memory on the information management server 3 (step 150). Next, among the combinations of the analysis key 106 and the purchase layer type 118 stored in the memory on the information management server 3, while there is a combination that has not been analyzed, Steps 154 to 1
Step 58 is repeated (Step 152). First, the category 131 specified in the memory on the information management server 3
The sales layer information is extracted from the sales information DB 14 for the specified purchase layer type 118 from among the sales results of the respective products belonging to and stored in the memory on the information management server. The sales information is extracted from the sales information DB 14 together with point-of-sale information indicating the time of sale (the purchase unit 115 may be set to indicate the time),
It is stored in the memory on the information management server 3. Each product code 1 of sales information on the memory on the information management server 3
Analysis key 10 for product code 104 corresponding to 16
6 and the scale 108 are extracted from the product information DB 12 and stored in the memory on the information management server 3. Each product code 116 of each sales record on the memory on the information management server 3
And the scale 10 of the analysis key 106 of the product code 104
8 is stored in the memory on the information management server 3 after being associated (step 154). Next, from the sales results including the point-of-sale information stored in the memory on the information management server 3 and associated with the scale 108 of each product, it is designated to be used when analyzing in chronological order. The sales performance of the scale 108 of the analysis key 106 or the attribute of the purchase layer type 118 specified to be used when analyzing in chronological order is extracted, and a preset boundary for totaling is calculated. After totaling for each time cycle represented,
The information management server 3 performs cross-tabulation for each scale 108 of the analysis key 106 or for each attribute of the purchase layer type 118, which is not designated to be used when analyzing in time series.
It is stored in the upper memory (step 156). Next, the time cycle of the scale 108 of the analysis key 106 specified to be used when analyzing in chronological order, and the purchase layer type 118 attribute not specified to be used when analyzing in chronological order It is tested by the chi-square test whether or not there is a relationship between sales trends. Alternatively, the time cycle of the attribute of the purchase class type 118 specified to be used when analyzing in chronological order, and the scale 108 of the analysis key 106 not specified to be used in analyzing in chronological order It is tested by the chi-square test whether or not there is a relationship between sales trends. The test result is stored in the memory on the information management server 3 and displayed on the information analysis client 4 (step 158).

[0025]

As described above, according to the present invention, by combining the sales information including the purchase layer information and the product information and analyzing the sales situation, the characteristics of the product and the product can be determined. The purchase behavior can be grasped by combining the characteristics of the customers who bought the product.

[Brief description of the drawings]

FIG. 1 is a configuration diagram of a network according to an embodiment.

FIG. 2 is a configuration diagram of a storage device on an information management server 3 according to the embodiment.

FIG. 3 is a PAD showing a procedure for registering product information of the embodiment.
FIG.

FIG. 4 is a diagram illustrating a data configuration of a product information DB 12 according to the embodiment.

FIG. 5 is a PAD showing a procedure for registering sales information according to the embodiment;
FIG.

FIG. 6 is a diagram illustrating a data configuration of a sales information DB according to the embodiment.

FIG. 7 is a PAD diagram showing a procedure for registering category information according to the embodiment.

FIG. 8 is a diagram illustrating a data configuration of a category information DB 16 according to the embodiment.

FIG. 9 is a PAD diagram showing a procedure for determining whether a significant characteristic is recognized in the sales situation by the purchaser by combining the product information and the sales information of the embodiment.

FIG. 10 is a diagram showing data developed on a memory in an actual mode.

FIG. 11 is a diagram showing data developed on a memory in an actual mode.

FIG. 12 is a diagram showing an example of a category, analysis key, and purchase layer type designation screen.

[Explanation of symbols]

3: information management server, 12: product information DB, 14: sales information DB, 16: category information DB

Continuing from the front page (72) Inventor Yukiko Fujiwara 890 Kashimada, Saiwai-ku, Kawasaki-shi, Kanagawa F-term in the Business Solution Development Headquarters, Hitachi, Ltd. JJ55 KK57 LL09

Claims (3)

[Claims] A product information provider analyzes information on attributes of a product indicating characteristics of the product, and a sales information provider analyzes sales performance including purchase layer information on a purchased customer by an information analyst. Each category, which is a group of product groups, is registered, the registration information is managed, and the sales key of the category is analyzed and the analysis key is used to analyze the sales performance, and the attribute of the customer who purchased each sales performance is displayed. A method of detecting that a significant feature is recognized in the composition ratio of sales performance in a system for cross-tabulation by purchase layer type. 2. The method according to claim 1, wherein when the analysis key or an arbitrary purchase layer type is arbitrarily selected, a product included in the category is selected from registered information of the analysis key or the purchase layer type. How to extract possible analysis keys or any purchasing demographics. 3. The method according to claim 1, wherein the composition ratio of the sales results is determined by an arbitrary analysis key or an arbitrary purchase layer type.
A method of detecting significant changes over time.