WO2002031721A1 - Purchase motive extracting method, fit commodity/fit customer extracting method, and device therefor - Google Patents

Abstract

A group of commodities fitted to the purchase motives of a customer according to the past purchase record of the customer is extracted, and information useful for selection of commodity by the customer is provided. A device comprises edge map storage means (600) where stored is an edge map that contains motives deciding purchase of each type of commodity and which has a same-kind multiple-type commodity list as a parameter and a motive candidate list including motive candidates potentially being motives deciding purchase of a specific type of commodity from the same kind of commodities as another parameter, purchased commodity list storage means (612) where stored is a purchased commodity list of each customer, means (614) for calculating a customer-edge correspondence map by accumulating purchase motives for each commodity type of the purchased commodity list, means (604) for searching for purchase motives of an unpurchased commodity type of a specific customer by using motives contained in the customer edge correspondence map of the specific customer as search keys. The commodity types sharing the common purchase motives contained in the customer-edge correspondence map of the specific customer are extracted.

Description

 Description Purchasing KM localization method and compatible products; Conformity provision method and equipment Technical field

 The present invention relates to a method for converting a purchase result t ^ of a purchase to a future sale.

'Eye〕

 Products that are often overlooked, such as music, music CDs, computer game software, vehicles such as translators, and customers select and purchase a particular view from a group of similar products. The merchandise is extremely numerous. “Customers often find it difficult to fit their self. Retailers are often troubled with the type of purchase and the appearance of the purchase. Product suppliers often have trouble predicting what their customers want in a product view, making it difficult to find products. In addition, a variety of marketing leaks are being developed for the future by analyzing purchased fibers, which are also difficult to set development goals. In the technology that seems to be the most advanced, the analysis of ίϋ¾'s “customer-purchased purchases of merchant outlets” using “customer attributes (age, mk annual income, etc.)” as a key, and “ Attribute-Seller that is often purchased by the customer of that attribute! ^ Customer group 'data is obtained, and the customer is presented with a group of merchant clients corresponding to the attribute to which the customer belongs, to assist in product selection. And provide the sales ^ s with an attribute that favors a particular business unit and display the auxiliary information for customer interception to develop sales © i activities. It is possible to provide auxiliary information for the production quantity and the development target by referring to the data of the attribute of the customer—the product type group frequently purchased by the customer of the attribute—the past purchase ¾a ’. Disclosure of the invention

However, the analysis of purchases by Okikura is based only on the attributes of customers as keys, and does not go deeper into customers. For this reason, analysis and M raw force ⁵ of the information obtained by ί氐Ku, for example, the customer is "customer attributes - instruments that are more purchase to the customer of that attribute When a customer believes and believes in 1 «¾ quality of view prepared based on the result of the analysis of the“ Housou group, ”the customer often satisfies the product ^! Δ.

 This invention #W is a technology that enables analysis that reflects the individuality of each customer, instead of ^!

 For example, consider computer game software. Fiber, there are about 100,000 buyers in Japan, and about 100,000 products are manufactured. In the past, there were about 5 purchases per capita, and there were 500,000 data for "customer-purchase merchandise. There are data for each customer. Analyzing this while reflecting each person's individuality says that Although it is easy to do, it is hard to do, and I do not know what to do.

 The customer purchases the product ABC and the customer buys the product: DEF. When there are 100,000 purchase bells' 髒 g, reflecting the difference between the customer and the customer, Useful information, such as recommending a product that matches the individuality to the customer who purchased the product ABC, or telling the retailer to specify the new product G based on the personality of the retailer's customer group I don't know how to get useful information to determine the amount of input.

 In the present invention, the above information, that is, the customer purchases the product ABC and the customer purchases the product DEF, and converts the past sales performance information into useful information while reflecting the difference in the customer's life I do.

 For this purpose, the present invention uses the edge theory. The edge theory has been developed by the present inventors, and can be roughly described as follows.

(1) There are a limited number of factors that customers can weigh in determining the size of a particular product from similar products with multiple locations, such as computer game software and m-passenger cars. The list of factors can be restored. For example, computer game software: ^ r, i buy from "no luck in magazines", i buy from "struggle: stimulating ΦΙ", i buy from "low« "," maker power " Product sales professionals who are in contact with many products and many customers, such as purchasing from `` can do it '', can list almost all factors that customers weigh with the fiber that appears to buy and sell. it can. For example, there are 57 factors in computer game software; ^, and almost all customers are listening to 57 types of factors, and have been told that they will not be able to view a particular view. Jageri refers to the group of factors that are evaluated by the fiber that the customer selects for specialty purchase as the edge group. (2) Individual quotients «have purchase motives» t in a group, have several purchase motives, and when purchasing ¾ ϋ ϋ 导 リ ス ト リ ス ト ϋ 専 門 専 門 商品, And the purchase of the product can be specified.

 For example, for product 、, there are customers who purchase from “high wisteria in ships, etc.” and customers who purchase from “stimulated struggle”, customers who purchase because of “low ¾”, and “ Can be accurately specified that there are few customers to purchase from. In addition, a data base that stores I », which represents the purchase of the merchant mouth, in association with the merchant mouth πί heavy is called an edge map.

 (3) The above information is provided, for example, by a wealth of physicians who have a lot of knowledge necessary to identify the cause from the patient's symptoms. It is almost in line with the fact that the preparation of medical tools will enable the same level of diagnosis to be made for doctors with “shallow physicians” or “deep physicians”, which is equivalent to the note data pace of product sales.

 (4) Once the edge map has been constructed, various concealments are possible. For example, a customer who has purchased a product ABC in the past may have a personality that prefers products that are “highly valued” or “stimulated instinct to fight” from the edge map, While it is expected that Jg is not recommended when recommended, it has a personality that does not fibrate "woven" or "manufacturer", and it is clear that selecting products from that perspective is also satisfactory. Alternatively, it can be seen that the new product has the P, q, r, etc. of the purchase, and that the new customers X have strong luck with the rinsing customers C;

 In the present invention, a useful theory is created by embodying the page theory in the following method or apparatus.

 (Claim 1)

The apparatus has a pager storage means 100 as shown in FIG. In the edge map 102 stored here, one parameter is set as a list of similar polysaccharide products. Ο In Fig. 1, the horizontal axis indicates the parameter set as Ochihachi, Β · 'as the parameter set. ing. The thick arrow in the figure indicates that the same kind of data continues. Another parameter in the edge map 102 is a list of groups of movements that can be used to recite the purchase of a particular product from a group of such products. In Figure 1: ^ ", the vertical axis shows the purchase tree ^ group a, b-' I have. Here, for example, a is “really high in etc.”, b is “stimulates the instinct of struggle”, c is “low price”, and d is “high in manufacturer”.

 The page map 102 states that each product customer purchases that product. In Figure 1, i½, the business port ^ «ΙΑ, is an example of the case where a is purchased because of a and sometimes g because of purchase of g, but it is almost always purchased from b to f. The device of 纖 纖 纖 1 、 、 装置 装置 装置 装置 装置 装置 装置

 Fig. 1 shows an example of a group of products that share mm (this ^ a and g). I have. The lotus F that shares i, a and g, the view DE that shares a, and the G that shares g are extracted. In addition, the example shows that the ACE that shares two purchase motives cf out of the purchases (here: ^ a b c f g) was extracted for the ceremony F. 110 illustrates an example in which a merchandise item ADEF having a is extracted for a purchase a, and a merchandise C F is extracted for a purchase cf (A Νϋ condition).

 According to the quotation, similar «products are similar in purchasing ®J¾! ^ Heavy groups can be extracted.

 Attempting to do so may reveal that completely unexpected merchants actually have very similar purchases. For example, two different game softwares of different genres, “Golf for Everyone 2” and “Dance Dance Revolution,” are “a maker that can be used by a friend”, “recommended by a friend,” It turns out that they share many purchases because they can play against them, because they are close to the game, and because they are likely to be useful. Many of the similar products found in this way are actually purchased by many customers. It is also recommended that customers who purchase one and not purchase the other be fully loaded with unpurchased products. In other words, it is emphasized that purchasing merchandise H is a similar product, in other words, it is a single merchant that is similar under the f @ I property of the purchasing customer.

When similar products are extracted in this way, it is possible to set up a product shop that is highly satisfying to customers, and to provide retailers with a useful “hidden store” that can be used to make accurate purchases. For the product supplier, we have useful 'sir β and H¾ targets to help us to increase or decrease the sales volume! ^ It is possible to provide useful data to extract similar data of 販 sales ^ ί © 1 activities. is there. (Method of Claim 2)

 This method will be renewed with the equipment shown in Fig. 2 and shown in Fig. 2. In this method, the edge map 102 is constructed, and the content of the edge map 102 is determined by the purchase of a particular product. Then, a group of similar merchandise outlets 208 that are similar in purchasing and power are extracted from among many similar products. The conditions for defining the edge map are the same as those of the device of claim 1: ^, and the description is omitted below. (Equipment of claim 3)

 As shown in FIG. 3, this device also has an edge map storage means 300 that stores the edge map 302, as in the device of the enclosure 1 shown in FIG.

 This device has a storage means 303 of an emotional personality map 305 which stores a special purchase fiber in a purchase sickle group corresponding to a personality at the time of purchase. This differs from the Contract 1 equipment. For example, there are people who buy game software overnight, ビ, and others who purchase as a character's fan, and others who play game ί¾ in the front and make purchases.

 In order to extract similar products from the group of customers who are often paid or purchased as a fan, there is not much similarity in purchasing motivation when testing game skills is the main subject. No problem. Therefore, the present invention has a memory means 303 of the emotional personality map 305 which stores the special purchase of the purchase group β in correspondence with the personality at the time of purchase. is there. In addition, the same person purchases as a fan, and the game ¾β comes to the forefront to make purchases, and there is a personality in the same person.

 In this device, the stored contents of the edge-muff storage means 300 are stored in the memory 302, and the means 304 for browsing using the purchase corresponding to the case as a key. For example, a group of merchants with purchase motives such as “Personality who appears to purchase as a fan, or personality who tries to expand the game in the foreground” is extracted.

For this reason, similar products when purchased as a fan, or suitable for their personality The corresponding product view is extracted and more useful information is obtained.

 (Method of Claim 4)

This method is typically performed using the device of the fiber 3 shown in FIG. In this method, the edge map 3 0 2 is constructed 4 000, and the emotional personality map that takes into account the purchase motives ^ f in the group f corresponding to the personality at the time of purchase is considered. Build 3 0 5 ¾ 4 0 3 and the contents of the edge map 3 0 2 ^ Running ⁴ 0 4 to search for as keys, to extract the goods «group 4 0 8 with purchasing« to M to rank among the likes of various goods.

 (Equipment of claim 5)

 This device is a device that manifests a purchase 1 item that a customer performs when selecting a product. As shown in FIG. 5, the device stores an edge map 13 'consideration means 500 and a purchase product list for each customer. Means for storing the purchased product list 5.12 and means for accumulating the purchase motives corresponding to the products included in the purchased product list for each customer 5 14 Power.

 According to this device, for each customer, the purchase SJW to be processed at the time of selection of one item or the customer is revealed from the past purchase responsibilities and output.

 (Equipment of claim 6)

 This device is a device that extracts and presents a business customer expected to be unsatisfied with the customer to a customer who is uncertain about product selection. As shown in FIG. 6, the edge map storage means 60 0, a purchase merchandise list storage means 6 1 2 that stores a purchase merchandise list for each customer, and a purchase page corresponding to a quotient lot included in the purchase merchandise list for each customer. The means 6 14 for calculating the correspondence map 6 16 and the unpurchased merchant rolls of the special H customers 購 Purchasing of the I group S) W are converted to the purchases included in the customer one edge correspondence map 6 16 A product (applicable product of the customer 6 18) which has the means of rising 6104 and shares the purchase included in the customer page correspondence map 6 16 of the customer is extracted.

 As described in the section 5 of the paper, the customer-edge correspondence map 6 16 shows the purchase SJ that the customer performs when selecting a product. This is based on past sales humility and is objective.

With this device, the purchase of an unpurchased lolo heavy group by a special customer Since the customer has a means of searching using the dynamic »contained in the correspondence map 6 16 as a key, the customer has a purchase fiber similar to the purchase of the product purchased by the customer in 3i. @ Types are extracted. The quotient extracted in this way is a product that is expected to have a high S for the customer. In fact, it is said that the products extracted in this way have high manpower.

 (Method of Claim 7)

 This method is shown in FIG. 7 in percentage terms; In this method, an edge map 600 is constructed 700, a purchase commodity list is compiled for each customer 712, and a purchase fiber corresponding to a commercial class included in the purchase commodity list for each customer. To build the customer-edge correspondence map 6 16: ¾ 7 14 and the purchase of the unpurchased goods group of 11 special customers are included in the customer's customer-edge correspondence map 6 16 Purchasing as a key: Eg704 is added. By this method, merchandise items sharing the purchase j ^ included in the customer page correspondence map 616 of the customer are extracted.

 (Claim 8)

 This device extracts the quotient response that is expected to be satisfactory to the customer who is uncertain about the product selection. In this respect, it corresponds to the device of the woven fiber enclosure 6. However, the device of the fiber wrapping system 8 responds to the fact that more than one person is involved in the same person, and when the customer's specific personality ffflB ffflB, it is expected that the customer will not be fully loaded. Extract the kind.

 As shown in Fig. 8, this device is included in the edge purchase list, the purchase item list that stores the purchase item list for each customer, and the customer purchase item list. A method for constructing a customer-to-edge map 8 1 6 by fibrillating the purchase process corresponding to the product view 8 1 4 and the features of the purchase scythe group according to the personality at the time of purchase Emotional personality Matsuf who is taking note of it. iai means 803 and special customers' unpurchased merchandise | weight group purchasing Ιίί 、 、 顧客 顧客 顧客 購 に 含 ま に 購 購 に にCorresponding emotional personality memory Means 8 0 3 Purchasing which is considered 3 Si has a means 8104 that dislikes Si »as a key, and purchasing i corresponding to the customer's specialty i ) Extract products 8 0 that fit ¾.

According to this device, when a customer is respected from a specific personality, a business outlet with a high degree of satisfaction is Is extracted.

 (Method of Claim 9)

 This method is renewed primarily with the device of the fiber 8 shown in FIG. In this method, 900 to build a page map and a list of purchased products for each customer are compiled: C¾912, and purchases corresponding to business 1 included in the purchased product list for each customer are accumulated. Build a customer-edge correspondence map 8 1 6 and build an emotional personality map that records specific purchase motivation in the purchasing group according to the personality at the time of purchase. The ES900 and the unpurchased goods of the special customer ¾ | group purchase are associated with the ranks of the customer included in the customer page correspondence map 8 16. By executing the marriage 04 using the purchase ®f »in the motion personality map as a key, a product type 920 matching the purchase ¾ ^ corresponding to the specific personality of the customer is extracted.

 (Apparatus of claim 10)

 This is a device that extracts customers who will have a high level of satisfaction with new products or @JW products. This device enables new efficient sales activities.

 As shown in FIG. 10, this apparatus has an edge map 3ft means 100, a purchased goods list storage means 1002 for storing a purchased goods list for each customer, and a purchased goods list for each customer. The means for calculating the customer page correspondence map 1 0 16 by making purchases corresponding to the product images included in the list and the purchase behavior of the unpurchased customer group of the special product «1 0 2 2» The means of searching for the purchase of specialty products in Hejiang 1102 2 »is used as a key 100 0 4 The customers that match the purchase incentives of the specialty type 1 0 2 2 1 0 2 4 Is extracted.

 According to this device, customers who will have a high degree of satisfaction with the product or special product intercepted from the W product are extracted.

 (Method of Claim 1 1)

This method is carried out on the apparatus of the fiber enclosure 10 shown in FIG. This method consists of accumulating 1 1 0 0 that constructs an edge map, n 2 1 1 1 2 that fibres the purchase merchandise list for each customer, and purchasing fibres corresponding to the merchandise cards included in the purchase merchandise list for each customer. Then, the customer-page correspondence map is calculated. ^ L 1 14 and the loyalty rule «Purchasing magic of the unpurchased customer group of 1 1 2 2» Use «as a key Perform ^ Ml 104. As a result, customers 1 1 2 4 conforming to the purchase of the specific merchant account ^D c «l 1 2 2 are generated.

 (Apparatus according to claims 1 and 2)

 This device extracts customers who will have a high degree of satisfaction with new products or ¾) «products. This device extracts the customers who will have a high degree of satisfaction with a single personality, in response to the fact that there is also a personality of the same person.

With this device, efficient sales activities can be performed.

 As shown in Fig. 12, this device is composed of an edge map, a storage means, a purchase goods list taking into account a purchase goods list for each customer, and a purchase goods list for each customer. Purchasing corresponding to the products included in the list! It describes how to build a customer-edge-compatible map 1 2 1 6 by using ^ 1 2 1 4 and a special purchase magic machine in the Purchasing Egoura group corresponding to the personality at the time of purchase. The emotional personality map is stored by associating the emotional personality map storage means 1 203 with the purchase 1 of the unpurchased customer group of the special type 1 2 2 2 with the personality conforming to the specialty 1 2 2 2. It has a means 1 2 1 4 to marry the purchase behavior »recorded in the three measures as a key. This device extracts customers 1 2 2 4 who have personalities that match the purchase motive of the 1 2 2 2.

 (Method of Claim 13)

 This method is executed by the apparatus of the fiber enclosure 12 shown in FIG.

This method builds an edge map: ¾ 1300, collects a list of purchased items for each customer ® 1 312, and purchase behaviors corresponding to the quotients contained in the list of purchased items for each customer. I 3 1 4 that accumulates machines and builds a customer page correspondence map, and an emotional personality map that takes into account the purchase of special knowledge in the purchasing sickle group according to the personality at the time of purchase Build a map: ^ 13 0 3 and the purchase of a group of unpurchased customers of a specific product 1 32 2, the specific business outlet.ェ モ 3 と し て 対 応 と し て ェ ェ «ェ ェ ェ 購 ェ ェ ェ ェ 購 ェ 購 ェ 購 購 購 購 購 購 ェ ェ 購 ェ 購 購 購 購 購 購 購 購 購 購 購 購 購 購 購 購 購 購 購 購 購 ェ ェ 購 購 購 購 実 行 購. By this method, customers 1 3 2 4 having a personality that matches the purchase motivation of the feature ^α 1 2 3 are extracted. BRIEF DESCRIPTION OF THE FIGURES FIG. 1: Explanatory drawing of the device described in claim 1.

Figure 2: Illustration of the method described in claim 2.

Figure 3: Illustration of the device according to claim 3.

Figure 4: Illustration of the method according to claim 4.

FIG. 5 _: Explanatory drawing of the device described in claim 5.

Figure 6: Illustration of the device according to claim 6.

Figure 7: Illustration of the method according to claim 7.

Figure 8: Illustration of the device according to claim 8.

Figure 9: Illustration of the method according to claim 9.

FIG. 10: Explanatory drawing of the device described in claim 10.

Figure 11: Illustration of the method described in Claim 11.

FIG. 12: Explanatory drawing of the device described in claim 12.

Figure 13: Illustration of the method described in Claim 13.

Figure 14: Diagram showing the system configuration of the device of the present invention.

Fig. 15 (A) to (C): Explanatory diagrams of edge maps.

Figure 16: An example of the result of ^^ edge map with purchase as the key.

Figure 17: Diagram showing an example of the emotional personality map.

Fig. 18: A diagram showing an example of a product similar to product F when viewed from personality.

Figure 19: A diagram illustrating an example of purchase product type data for each customer.

Fig. 20: Expanded purchases corresponding to merchandise weights included in the purchased merchandise list for each customer. Fig. 21: Diagram of cumulative purchases shown in Fig. 20 for each customer.

Fig. 22 (A) and (B): Diagrams of finding and finding customers that fit new product X.

Fig. 23 (Α) and (Β): Diagram showing extraction of products that will satisfy customer 満 足. Fig. 24: A diagram illustrating the customer's personality through the customer-edge correspondence map shown in Fig. 22 (Β).

Figure 25: Personality: Figure showing how κ comes to the forefront and selects products suitable for customers who return products.

Figure 26: Diagram showing an example of a page map. Figure 27 _: Diagram showing a list of purchased products by customer.

Figure 28: Diagram showing customer-to-edge correspondence maps obtained from Figures 26 and 27.

Figure 29: Diagram showing the details of how to obtain Figure 28.

FIG. 30 (A) to (C): Diagrams showing how a product suitable for the customer is extracted from the mapped customer-edge correspondence map.

Figure 31: Diagram showing how to extract products that fit customers from the cft-converted customer-to-edge correspondence map.

Fig. 32 (A) to (C): Diagrams showing how to extract customers that fit the product.

Fig. 3 3: Diagram showing how to identify customers that match the product.

Figure 3 4: Product List by Genre 1

Figure 35: Product List by Genre No.2

Figure 36: Upper row: Similar product groups extracted by edge theory and cMs where buyers of specific products are purchasing similar products. Lower: Purchasing products belonging to the same genre ¾ Indicates ο

Figure 37: Personality is identified from the customer-edge correspondence map, and Table 1 is a table that summarizes whether or not products corresponding to the identified personality are actually purchased.

Figure 38. Personality is identified from the customer-to-edge correspondence map, and Table 2 shows whether the product corresponding to the identified personality is actually purchased.

Figure 39: Personality is identified from the customer-edge correspondence map, and whether or not the product corresponding to the identified personality is actually purchased is part 3 of the table.

Figure 40: Personality is identified from the customer-to-edge correspondence map, and Table 4 shows whether the product corresponding to the identified personality is actually purchased or not.

Figure 41 1: Personality is identified from the customer page correspondence map, and Table 5 shows whether the product corresponding to the identified personality is actually purchased or not. Best practice for carrying out the invention

FIG. 14 shows a system configuration of the apparatus of the present invention, in which a small product for selling a product (the product is not limited to a tangible object, but includes a woman's product, that is, for example, various periods. Deposits, etc. are similar products mentioned here, and time deposits for each period are It is a product of a lot of goods, and the store of ts 亍 is a small part.), An analysis center that has many small items, a product supplier, a sales promotion activity, and a customer are involved. Each subject is equipped with a device that is used at the No. 1 personal convenience store, a small device is provided with a small information device 1402, and a product supplier is provided with a product supplier device 1404. Provided, the sales promotion source has sales promotion source information mm i 406, the customer has a customer's planning device (sometimes a mobile sis compatible with ι mode), and the analysis center has an analysis. It is equipped with Senyuki Yuki device 1 4 1 2. All of the Oki g¾a devices are started by a computer network typified by a computer network typified by In-Yu-Ichi-Net, which constitutes each weight device of the present invention as a whole ^ s' mmi 4 o 2 is connected to the in-store regis- ter, so that every time a customer purchases, which customer purchased which product «! Enter ^ to send to the analysis center.

 Analytical Sensing "If the device 1 4 1 2 collects the" t * f "of which customer purchased which type from a small customer by customer and" customer-to-purchase product type list 1 4 1 Create 4 ”. While most retailers subscribe to the system in Figure 14 while customers may purchase at different churn, aggregating at the analysis center will create a near-customer-to-purchase product plan. List 1 4 1 4 "is put together.

The analysis sensor device 1 4 1 2 is equipped with an extract base (hereinafter, referred to as D / B) 1 16 from the two factories. One is the age map 14 18 and the other is the emotional personality map 14 20. As shown in Fig. 15 (B), the page map 1 4 18 shows one parameter set as a similar multi-lingual list and another parameter set as a specific restaurant from a group of similar products. A group list that can be used to purchase the merchandise of a particular type, and a two-dimensional map that uses the two types of parameters is used to record the purchases of that type of merchandise for each type of merchandise. I have. As shown in Fig. 17, the emotional personality map 1440 corresponds to the personality at the time of purchase, and the special character [@buy! It is marked with ^. The Edge Map 1 4 1 8 and the Emotional Personality Map 1 4 2 0 are prepared by a sales 51 * who has deep product knowledge and customer knowledge in contact with many products and many customers, and a deep doctor This is equivalent to an expert who performed MS on the relationship between symptoms and causes. The analysis device "^ 1" device 1 4 1 2 creates a similar product image group list 1 4 2 4 based on the edge map 1 4 1 8 based on the purchase map. Based on 1 4 1 8 and the emotional personality map 1 4 2 0, create a product group list 1 4 2 6 that matches one personality that appears at the time of purchase. The customer creates an analysis list 1 4 2 8 of the purchase orders that the customer processes at the time of purchase.Furthermore, from this Purchasing by customer »List 1 4 2 8, the customer list 1 4 3 0 that matches the product and the Create a list of suitable products 1 4 3 2. Furthermore, create various types of products 1 4 3 4, for example, a version that requires a customer's fiber that fits a certain product. The data is according to customer's request, ¾fl® device 1408, small '«1 device 1402, goods supply! ^' Suiyuki ¾0« place 1404, foot Domoto 'Kiyoyuki «sent equipment to 1 4 0 6. This ^", information that is not sent information to be transmitted in accordance with enough demand of consideration to the protection of the individual is defined go-between by the pre-lawyers. In addition, plows and cranes are used to pay for data, etc. Communication between devices uses comfort technology to ensure high information security.

 Hereinafter, the apparatus and method of the present invention will be described based on an example in which computer game software (hereinafter referred to as game software) is sold.

 Fig. 15 is a diagram illustrating an edge map that is an exotic D / B. The game software AB C · · It is often purchased or hardly purchased depending on the fiber.

It is extremely difficult to predict whether a product will or will not be accepted in the marketplace, which has plagued small-office purchasers. It is also difficult for manufacturers to determine production. Many market methods have been developed ^ For example, the data of “customer attributes — quotients π ^α β! Selling a lot of merchandise purchased in large quantities {© 1 enables efficient sales ^ 販 © 1 Activities can be developed, but accuracy can be applied to new products In fact, the purchaser appears to be in the hands of the purchaser's intuition. Therefore, the intuition is often lost, and there are many missing items. Also, while manufacturers are making great efforts to develop products that are acceptable to the market, cows often have to stay out of the way to be accepted in the market, which can lead to disappointment. Many. In addition, it is difficult for customers to select β that satisfies them, and they often select mtti products.

 The inventors of the present invention have repeated the I-mochi using the above-described technique for making opacity possible, and have finally arrived at the page theory. The edge theory has been developed by the present inventors and can be roughly described as follows.

 (1) There are a finite number of factors that customers consider when selecting a particular product from similar products, such as game software, and a product sales expert can restore the factor group. For example, purchase from game software ゲ ー ム ^ r, “Thank you very much for ¾ϋ¾, etc.” ^, Purchase from “Struggle: stimulated” ί, purchase from “low price”: ¾ ^, “Manufacturer Buy from "friendly": ^ · There are many products and many customers in contact with the product sales specialists, the list of factors that the customer weighs with the fiber to watch the purchase is almost complete can do. There are 57 factors for computer game software, and almost all customers choose a specialty restaurant after considering 57 factors. According to Edge Science, a group of factors that a customer weighs at a wisteria that intersects a particular customer is called a purchase sickle group. On the vertical axis of Fig. 15 (、), the dynamics that make it difficult to purchase a specific type of product from a certain type of product group are taken as a parameter. ing. In addition, the horizontal axis uses a multi-product list ABC C '' as a parameter.

(2) Individual product views (for example, A)

It has several purchases in c · · ', and when a purchase group list is entered, a product sales expert can identify j »for each product to purchase that product. For example, for product A, there are customers who purchase from the purchase fiber a “highly debated” etc. and customers who purchase from the purchase g “can compete with someone”, and purchase “c” It is possible to accurately specify that there are almost no customers who purchase from the “price” or purchasing customers from “manufacturers of the purchased fiber” (Fig. 15 (B)). The list AB C ■ 'is used as a parameter, and the vertical axis is a two-dimensional map that uses the purchase sickle group list abc ·-• as a parameter. An edge map that stores a purchase request to purchase a product view is indicated by a thick arrow, indicating that the same type of information is more ubiquitous.

(3) Once the edge map shown in Fig. 15 (Β) is constructed, Management becomes possible. For example, a customer who purchased a product ABC would have a personality that prefers products that are “highly valued by HS” or “compete against someone” from the edge map, and leaks a product with those purchases. It can be expected that you will be satisfied with J, but you also have a personality that does not emphasize “« ”and“ manufacturer ”, and you can see that it is satisfactory even if you select a product from | ¾¾

 First, a technique for extracting merchant-mouth horoscopes with similar purchases from the edge map illustrated in Fig. 15 (B) will be described.

 An edge map shown in FIG. 15 (B) can be obtained by picking up a purchase! ^ For each product by a product sales expert. This map will expand as new products are introduced.

 Once this edge mapping is performed, a variety of useful data can be obtained. Extracting business groups with similar purchase motives. There are two methods. The first method is to specify a product and dislike products with purchased fibers that are similar to the purchase of the product. Figure 15 (C) illustrates the results of the extraction. For example, product A has purchase motives a and g, and product F has the two purchase motives (degree of agreement 100% ), The product IDE G has one of the purchases (50% agreement), indicating that the product BC is not quite similar. Similarly, for product F with purchased fiber ab e f gh, product C shares three purchases (50% of the total), and the product of 0 fe is a single fiber. Two S¾ in the figure indicate common purchasing motivation (edge).

 For example, if it is found that product F has the same purchase as product A, various things become possible. For example, recommending product F to a customer who has product A but not product F can probably be expected to be satisfactory. Based on the sales performance of product A in the past, small can show the purchase amount of product F. Similarly, the sales activity plan of product F can be i ^ based on the degree of effectiveness of past sales and release activities of product A. Manufacturers can also make ^ S for product F with reference to past sales of product A. In addition, by intensively analyzing past hit products, it is possible to find a set of purchasing magic machines ^: common to hit products, and the maker will have clear development goals.

Figure 16 shows an example of the result of using an edge map with the purchased fiber set to Tsuru-Ki. By referring to this list, it is clear that the set of purchased items commonly shared by heat products ^: It becomes. For example, quality and F are heat products: It is understood that products that combine t and purchase cfh are accepted by many customers.

 Figure 17 shows the ^ (columns) of the emotional personality maps 30.5 and 1440. For example, buy game software: ^^. Game; Some people come to the forefront of trying KM.

 Personality indicates the purchase motivation that is weighed when purchasing as a fan. This ^ indicates that purchase ethics are important and purchase ethics are not weighed. The personality ^ indicates the personality of ¾ ^, who makes the purchase by trying out the game 試 す β. In this case, the purchase ab d e f is emphasized, and the procurement c is not considered.

Fig. 18 shows an example of a product similar to product F when viewed from personality (measures the purchase cfh), where rn ^ r and product C have 100% 1! ^. I understand. As shown in Fig. 15 (C), when comparing one product with another product without assuming a specific personality, product C has a 50% coincidence with product F, whereas personality a Return the product to the scale: ^ shows that Product C matches 100% of Product F. On the other hand, for products A and E, the matching degree of the merchant account ^Q cr ^ products in Fig. 15 is 33%, so there are some products that are eligible products. The degree of coincidence of ^ す る is 0%, which indicates that it is not actually a conforming product.

Fig. 19 shows an example of the "(" column of purchased goods for each customer. Fig. 20 shows the edge map shown in Fig. 15 (B) and the purchased goods list by customer shown in Fig. 19 Fig. 21 is an expanded view of the purchase 商 corresponding to the business unit ^Q c «included in the purchase merchandise list for each customer. Fig. 21 is a diagram showing the purchase edition of Fig. 20 for each customer. For example, it can be seen that the customer selects the product by ftf processing the purchased fiber ab fgh i and selects the product without paying attention to the purchase and c.

 The “purchase list by customer” in Fig. 21 is useful by itself, but as described below, it is a flat application for various applications.

Figure 22 shows that a customer who satisfies the new product X (or may be in stock, but the stock: ^ has not purchased the product yet! / Find Hi? (A) shows an edge map of the product X. (B) shows the purchase re A list, that is, a customer page correspondence map is shown. In (B), the edge corresponding to the edge map of the product X has a double circle. Of the six purchases that are important to the customer, three out of six purchases are satisfied with item X and the satisfaction is 50%, while the five purchases that are important to the customer are It can be seen that the four items are satisfied and the satisfaction is 80%. Obviously, Product X fits better with the customer than the customer. Figure 23 shows the selection of products that customer お そ ら く will probably be satisfied with from products that have not yet been purchased by customers B: B to F and X to Z. (A) shows a customer page correspondence map of the customer. (B) shows an edge map of an unpurchased product. In (B), the edge corresponding to the customer-to-edge map of customer ゥ is marked with a double circle. For products F and Y, up to five out of the six purchasing motivations that the product has are in agreement with the customer's purchased fiber, and the conformity is 83%. Obviously, it can be expected that the customer will have high satisfaction with product F and product Y.

 FIG. 24 explains the manner in which the customer's personality is ^! From the customer-to-edge map shown in FIG. 22B.

 As shown in Fig. 17, for example, it is known that the personality α is applied to five purchases of acgh i. In Figure 24: ^ ", from the customer-to-edge map, the customer has the personality /? In front of him to select a product, and the customer て has the personality / personality key to select the product. In other words, it can be seen that customer ゥ has a dual personality in selecting the product.Figure 25 shows that the personality is in front and extracts products that match the customer who intersects the product » Customers with personality can see that products C, F, Y, and Z are compatible.Comparing this with the results in Figure 23, it is clear that by introducing personality, Compatible products will be expanded (this ^^, it is newly found that product Z is a compliant product.) Also, it can be seen that the conformity is improved, and the compliant products can be more reliably extracted.

 Conversely, the product C FYZ is well suited to personality. Therefore, in order to extract customers that fit the product C FYZ, it can be seen that it is sufficient to extract customers who are known to have personality as shown in Figure 24.

Fig. 26 shows an example of converting the edge map into a ¾M. Until now, purchasing! In contrast to whether or not to plant the tree in ^, this time we will digitize the carrying of the bite. This figure is also made objectively by product sales experts. Figure 28 shows the edge map visualized in Figure 26 and the customer-to-edge map obtained from the purchased product list by customer in Figure 27. The details of how to obtain it are shown in Figure 29. It is simply an averaging, for example, a customer purchases a product A with an edge of a of 3 and a product B with an edge of a of 2; ^ It is calculated as having weight. In Figure 28, it has been transformed into a single building.

 Figures 30 and 31 show how to extract products suitable for customers from the quantified customer-to-edge map.

 Here, in addition to the β-converted data, heavy engineering is used. The heavy industry is a buyer who is particularly valued. For example, a customer i's purchased fiber a has a value of 1; i is calculated as 1 because one of the 10 purchased items happens to have a purchase motivation a of 10! For example, ^^ of the 10 purchased items may have a concession of 1 in the purchase a. In order to clarify this difference, in this example, a heavy industrial machine is used. In FIG. 30, the customer indicates that the purchased heavy fiber b is a heavy duty. -Goods also have heavy jobs, for example, most of the customers who purchase product B are purchased with the purchase »« b fti, so "b" is the heavy edge for product B.

 As shown in Fig. 31, products matching the customer are extracted. ¾ ^ First, products with matching heavy edges are preferentially found. This: t, the product BD G matches, and the product ACE F does not match. Next, the product with the highest number of matching pages is determined as the product with the matching degree. In the customer-edge map (A) and the edge map (B) in Figure 30, zero is a non-detailed purchase. Therefore, when comparing the customer-to-page map (A) and the edge map (B) in Fig. 30, the number of columns (the number of pages) that both have zero percent and a direct count are counted. For example, it can be seen that Product B shares four edges with the customer and Product D shares three edges with the customer.

 Have the same number of edges; ^, Furthermore, a product with a small deviation is regarded as having a low fitness. Here, the deviation refers to the sum of deviations of the corresponding edges other than the heavy engineering edge at the corresponding edge, and is exemplified in FIG. 30 (C). The column referred to as cross hatch in (C) corresponds to a column with no heavy industry or no page. No deviation is calculated in these columns.

The above logic, that is, (1) Products with matching heavy engineering pages have higher fitness than products without matching.

 (2) For products matching heavy engineering pages or between non-matching products, products with a large number of matching edges are regarded as having high fitness.

 (3) If there is still no difference between products, it is assumed that the product with a small deviation in Fig. 30 (C) has a high fitness.

 According to the supply of Fig. 30 and the transfer of Fig. 31, it can be seen that the products with higher relevance are GBD FEAC (AC is equal) in order.

 Products that have high relevance extracted in this way are actually accepted by customers with high satisfaction.

 With almost the same language, it is possible to extract customers that fit the product. First, as shown in Fig. 32 and Fig. 33, the customer whose heavy edge coincides is found first. In this case, the customer aka is matched and the customer is not matched. First, the former is regarded as a customer with a high degree of conformity. Next, the number of matching edges! Customers with high relevance. In Figure 32, the edge map (A) and the customer-to-edge correspondence map (B) show that the customer aka shares four edges with product B, the customer a has four edges, and the customer It can be seen that ゥ shares three edges.

 The number of shared edges is equal; ^ r, and customers with small deviations are judged as good fitness. The deviation is illustrated in Figure 32 (C).

 The above logic, that is,

 (1) Customers with matching heavy edges have higher fitness than customers without matching edges.

 (2) It is assumed that the number of matching pages is high between customers who match heavy industries or between customers who do not match each other, and the degree of conformity of the customers is high.

 (3) If there is still no difference between the customers, it is assumed that the customer with a small deviation in Fig. 32 (C) has a high fitness.

 According to the supply shown in Fig. 32 and the item shown in Fig. 33, it can be seen that the customers with higher relevance are Aka Oewi in order.

 The customers with high relevance extracted in this way are expected to accept the product with high L 高 satisfaction.

Figures 34 and 35 list game software products of the same variety β by genre. Things. Here, the genre is changed from A to W ^ ϋ. ^ S ^ I contains multiple items or items. Hereinafter, a specific product or view is indicated by a genre symbol and a Kato number. For example, the first product in the power battle, Digimon Field Digi-Like Battle, is denoted by A1.

 The upper part of Fig. 36 shows a group of similar products for which similarity was found by the edge theory to the product shown at the top. “1” indicates the number of common edges in the 72 edge groups. For example, a product of A1 and a product of D22 have 26 edges (showing that they share purchases. The figure shows the top 20 similar products with the highest number of matches. However, if there is a similar product with the same number of matches, ^ indicates the product of the same rank, and 20 or more similar products are listed: ^ According to page theory, For example, from a purchase perspective, it can be seen that product A 1 belonging to force battle is similar to product D 2 2 belonging to the genre of RPG and product H 1 belonging to simulation RPG.

 The view column shows the number of customers who have purchased similar products listed among the 1 1 2 people who purchased product A 1. For example, it indicates that 32 out of 1 1 2 people who purchased the product A 1 purchased the product D 2 2, and 7 people purchased the product H 1.

This one night is a five-year purchase record of a group of 18 game software stores and 105 members.

 The lower part of Fig. 36 shows purchased fibers by genre. For example, out of 1 1 2 people who purchased product A 1, a total of 77 customers purchased A 2 1 from other products A 2 belonging to the same genre. The upper and lower rows have the same number of products.

Clearly, if you change the total quantity up or down it, it will be clearer that the yu, br

(Of the 115 people who actually bought B1, only B7 ~: B22 owned by only 70 people), whereas similar products shown above were a total of 1 2 It can be seen that the property owned by one person is more likely to be satisfied when recommending this similar product.

 Fig. 37 and subsequent figures show the results of extracting personality by the method of Fig. 17 and extracting the products that match the personality by the method of Fig. 18 and the results of comparing actual purchases. Obviously, it is considered that the products extracted by the method in Fig. 18 are very high in the actual purchase. As a result, it can be seen that if an unpurchased product is included in the product group extracted by the method in FIG. 18, the product will be almost satisfied if the product is recommended. .

 The difficult modes described above are merely introductions of the columns of the present invention, and the scope of the present invention is not limited to the examples. Those skilled in the art can implement the present invention in various forms within the spirit of the claims.

 The present invention is based on the idea that the edge theory is used to convert the purchased fiber data into individual customers. In addition, the product supply can be beneficial to the main actors and the like, and can suppress β and muta produced at the »stage from product sales.

Claims

1 One parameter set is a list of many similar items, and the other parameter set is a list of groups that can be used as a fiber for purchasing a particular item from a group of such items. Edge Matsuf¾1 means that memorizes the product «Purchasing the magic J»

 ヅ 手段 ヅ ヅ 憶 ヅ ヅ ヅ ヅ ヅ 購 購 購 購 ¾¾ ¾¾ 購 X

 A device that extracts a group of products with similar purchases from a variety of similar products.

2—One paramete list is a list of similar and diverse products, and another paramete is a list of groups that can be used to show the purchase of a particular item from a group of such products. Build an edge map that allows you to make purchases of that product type:

 A step of searching for the contents of the page in the purchase of the goods of the particular model by using

 A method of extracting a product type group with similar purchases from a variety of similar products.

3 One parameter set is a list of similar and diverse products, and the other parameter set is a list of Satsuma groups that can be used to purchase products of a particular customer from a group of such products.

An edge map sti means taking into account the purchase of «the quota Lolo c« for each «

 According to the personality of the purchase ¾, the emotional personality map that remembers the specific purchase motivation in the purchase behavior «»

 EdgeMuff "13. Means of Hiragana, with means to use key purchases corresponding to special

A device that extracts a group of products with purchasing incentives that suit the case from a variety of similar products 4 One parameter set is a list of similar and various products, and the other parameter set is a list of scythes that can be used as a fiber to show the purchase of a particular product from a group of such products. Construct an edge map that says:

 According to the personality at the time of purchase, build a emotional personality map that describes the characteristics of the purchase group.

 The content of the edge map is supported by a key purchase request;

 A method of extracting a group of merchandise that has a specially purchased SJ from a variety of similar products.

5 One parameter set is a list of similar multi-products, and the other parameter set is a list of groups that can be used to purchase a particular item from a group of products. Etsujimuff remembers the purchase of the business account ^a c «every time,

 A purchase product list storage means for storing a purchase product list for each customer;

 It has means to accumulate purchases «corresponding to merchandise included in the purchased merchandise list for each customer,

 A device that outputs a purchase list for each customer.

6 One parameter set is a list of similar and diverse products, and the other parameter set is a list of groups that can be used as a fiber for purchasing specific products from a group of products. Each month, you have to make sure that you buy the product type.

 A purchased product list storage means for storing a purchased product list for each customer;

 Means for accumulating the purchases corresponding to the merchandise customers included in the purchase merchandise list for each customer and calculating a customer page correspondence map;

There is a means to mirror the purchase ijW of the special merchant's buying and selling mouth group of customers with the purchase ii »included in the customer's customer-edge map as a key, A device for extracting a business partner who shares a purchase included in a customer's customer service correspondence map.

7 One parameter set is a list of similar and various products, and the other parameter set is a list of groups that can be used to purchase products of a specific garment from a group of such products.

Build an edge map that says aw to buy the business unit every time:

 When you ® を the list of purchased products for each customer,

 By accumulating the purchases »corresponding to the commodities included in the purchase commodity list for each customer, and constructing a customer-edge correspondence map,

A non-purchase quotient opening ^{α βι} group of purchasing i »of special ¾¾ customers, running the customer included in the customer one edge corresponding Madzupu purchase Ιδ» as a key! ^ To ®,

 Those who extract business contacts that share purchases included in the customer's customer page correspondence map

8 One of the items, "Lame Ichiyu, is a list of many similar products, and the other, I. Lame Izu is a Satsuma that can be used as a feature to purchase products of a particular customer from a group of such products. An edge mapf that stores a group list and stores a cut for purchasing the product for each product;

 A purchased product list storage means for storing a purchased product list for each customer;

 Means for accumulating the purchase motivation corresponding to the commodity customer included in the customer purchase commodity list and constructing a customer page correspondence map;

 Motivational personality map that memorizes the special purchase motives in the gffi group according to the personality at the time of purchase.

 The special merchant mouth p «Group purchase 1» of the special 11 customers is stored in the emotional personality map Bft means in association with the specific personality of the purchase S «© included in the customer's customer correspondence map. Have a way to view purchasing as a key,

A device that extracts product types that have purchasing incentives that match the customer's rating. 9 One parameter set is a list of similar and diverse products, and the other parameter set is a list of Satsuma groups that can be used as a drop-off for purchasing products of a particular lotus from a group of such products. If you build a page map that describes the magic of purchasing the quotient in 11

 By listing the purchased products for each customer,

 By accumulating the purchase motivation corresponding to the commodity bacteria included in the purchase commodity list for each customer, a customer page correspondence map is constructed :! When,

 Build an emotional personality map that takes into account specific purchasing motives in purchasing behaviors, corresponding to the personality at the time of purchase.

 The purchases stored in the emotional personality map by associating the purchase s »of the special merchant's trade product iw with the rank of the purchase 1» included in the customer's customer correspondence map. Execute π¾ with »as the key, and

 A method of extracting merchandise customers with purchasing motivation that matches the customer's case.

1 0 One parameter is a list of similar bacteria and another parameter is a list of awake ^ i groups that can be used to purchase products of a specific page from a group of products of that type. Edge Matsufou remembers the purchase of the product view for each view,

 A purchased product list storage means for storing a purchased product list for each customer;

 Means for accumulating the purchase motivation corresponding to the products included in the purchase product list for each customer to calculate a customer page correspondence map;

 Unpurchased customers of special order ^ ω purchase prudent, its special order. Have the means to make regular purchases key,

 A device that extracts customers who meet the purchase motivation for that particular product type.

1 1 One parameter overnight can be a list of many similar products, and the other parameter overnight can be a purchase of a particular item from a group of such products. When a group map is created and a page map is stored that stores a SW for purchasing the product for each product, The process of を ^ of the purchase product list for each customer,

 By accumulating the purchases corresponding to the product images included in the purchase product list for each customer and calculating the customer page correspondence map,

The purchase of ^ Offer customer group of special ^ goods, the special ^ B ^D C »of purchasing 1» as a key;?! Running to ^,

 A method of extracting customers who meet the purchasing motivation of the special product type.

1 2 One parameter set is a similar long list, and the other parameter set is a list of potential groups that can purchase products of a particular type from a group of similar products. An edge-muff memory means that stores | ¾ »

 A purchase product list that stores a purchase product list for each customer.

 Products included in the purchased product list for each customer f Fuji :! Means for accumulating purchasing incentives corresponding to each customer and building a customer-page correspondence map;

 According to the personality at the time of the purchasing decision, the motivational personality 手段 which considers the specific purchasing motivation in the group

 It has a means to purchase a group of buyers of a specific product according to the personality that fits the particular product and use the purchase i ェ »as a key in the emotional personality map.

 A device that extracts customers who have the personality that matches the purchase of the special product.

1 3 One parameter overnight can be a list of many @@ items, and the other parameter can be a purchase of a particular product from a group of such products. Build a group list and build an ET map that remembers the purchase of the product for each product type ^,.

 Purchase product list for each customer! ^

 Accumulate the purchase motivation corresponding to the products included in the purchased product list for each customer to build a customer page correspondence map;

According to the personality at the time of purchase, the special purchase in the purchase S <i Build a emotional personality map :! ¾ and

 Purchasing an unpurchased customer group of specialty items »», and matching the personality that matches the specialty ^ @ | with the purchase | ¾ »key 311 described in the emotional personality map. hand,

 A method of extracting customers with personalities that match the purchase motivation of the special feature β.