Intelligent commodity and detail identification matching processing system and method based on big data

Abstract

The invention discloses an intelligent commodity and detail identification matching processing system and method based on big data, and relates to the technical field of commodity intelligent identification matching, and the method comprises the steps: evaluating the commodity deviation degree of a user for commodities in a commodity sales platform, and obtaining historical deviation commodities; acquiring commodity detail data of historical biased commodities, acquiring commodity attribute data from the commodity detail data, and analyzing attention degrees of users to different commodity attributes of commodity types to obtain key commodity data; obtaining a historical purchase record of a user, obtaining historical biased commodities, analyzing the influence degree of the matching state between different commodities on commodity sales, and obtaining feature matching commodity data; and obtaining biased commodity data, feature matching commodity data and a commodity sales platform of the user, carrying out intelligent commodity matching on the user based on the biased commodity data and the feature matching commodity data, and carrying out commodity recommendation on the user on the commodity sales platform.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent commodity identification and matching, and in particular to an intelligent commodity and detail identification and matching processing system and method based on big data. Background Art

[0002] At present, product recommendations and matching for users are mainly based on the analysis of users' historical behaviors and historical purchase records to obtain product interest data of users, and based on the product interest data, products are screened to find products that meet the requirements and recommend them to users. However, even if two products are of the same type, the product details information of the products may be very different. In addition, the conventional method analyzes users' historical behaviors and historical purchase records, and the product demand data obtained is not detailed enough. For two products of the same type, further screening cannot be performed. This will result in that the recommended products are not highly matched with users, which reduces users' desire to purchase products on the platform, affects users' purchasing decisions, and may even affect users' trust in the platform, damage the overall image of the platform, lead to the loss of many customers, and seriously affect the interests of the platform. Summary of the invention

[0003] The purpose of the present invention is to provide a system and method for intelligent commodity and detail identification and matching processing based on big data to solve the problems raised in the prior art.

[0004] To achieve the above-mentioned purpose, the present invention provides the following technical solution: a method for intelligent commodity and detail identification and matching based on big data, the method comprising:

[0005] Step S100: Obtain the user's historical browsing records and historical purchase records, evaluate the user's preference for commodities on the commodity sales platform, and obtain historically preferred commodities;

[0006] Step S200: obtaining the user's historical preference products, obtaining product detail data of the historical preference products, obtaining product attribute data from the product detail data, analyzing the user's attention to different product attributes of the product type, and obtaining key product data;

[0007] Step S300: Obtain the user's historical purchase records, obtain historical preferred products, analyze the impact of the matching status of different products on product sales, and obtain characteristic matching product data;

[0008] Step S400: Obtain the user's preferred product data and feature-matched product data. The product sales platform performs intelligent product matching for the user based on the preferred product data and feature-matched product data, and recommends products to the user on the product sales platform.

[0009] Furthermore, step S100 includes:

[0010] Step S101: Acquire each historical browsing record and each historical purchase record of the user;

[0011] Step S102: Obtain the total number α of each historical browsing record, obtain the total number β of each commodity purchase record, and calculate the characteristic coefficient γ=β / α of the historical browsing record;

[0012] Step S103: Extract the types of products browsed and purchased by the user from the historical browsing records and historical purchase records, and calculate the user's type preference for each product type, where the user's type preference for the a-th product type Q a =(C a ×γ+B a ) / (2×β), C a Indicates that the product type browsed by the user is the total number of historical browsing records of the a-th product type, B a Indicates the total number of historical purchase records of the a-th product type purchased by the user;

[0013] Step S104: When the type bias degree Q a When the value is greater than the preset mark type bias degree threshold, it is determined that the user has a bias towards the a-th item product type, and the a-th item product type is recorded as the user's biased product type;

[0014] Step S105: obtaining several preferred commodity types of the user, and when the commodity type of a commodity browsed by the user or a commodity purchased by the user is the user's preferred commodity type, the commodity is recorded as a historical preferred commodity.

[0015] Further, step S200 includes:

[0016] Step S201: Acquire each of the user's historical preferred products, acquire product detail data of the historical preferred products, and acquire product attribute data from the product detail data, wherein the product attribute data includes text information corresponding to several product attributes of the preferred product type of the historical preferred products;

[0017] Step S202: analyzing the user's attention to several product attributes in several preferred product types, wherein the specific analysis process of the user's attention to the e-th product attribute of the d-th preferred product type includes obtaining text information of the e-th product attribute in the historical preferred product of the d-th preferred product type, performing text preprocessing on the text information, and obtaining a product attribute keyword group;

[0018] Step S203: Obtain the total number N of several historical biased commodities of the commodity type of the dth biased commodity type d , calculate the characteristic attribute vector of each product attribute keyword in the product attribute keyword group of the e-th product attribute in the historical biased product, where the product type is, the characteristic attribute vector T of the λth product attribute keyword in the product attribute keyword group of the e-th product attribute in a historical biased product of the d-th biased product type λ :

[0019]

[0020] Among them, μ λ It is represented by the total number of occurrences of the λth product attribute keyword in the text information of the eth product attribute in a certain historically biased product; L λ It is represented by the total number of product attribute keywords in the product attribute keyword group of the e-th product attribute in a certain historically biased product; N d,λ It is represented as the total number of historically preferred products containing the λth product attribute keyword in the product attribute keyword group of the eth product attribute;

[0021] Step S204: obtaining a marked attribute vector of the e-th item of product attribute in the historically biased product based on the characteristic attribute vectors of each product attribute keyword in the product attribute keyword group of the e-th item of product attribute in the historically biased product;

[0022] Step S205: Based on the marked attribute vector, calculate the attribute details similarity U of the e-th item attribute between a certain historical biased product and several historical biased products e :

[0023]

[0024] Among them, S e It is represented as the label attribute vector of the e-th commodity attribute in a certain historical bias commodity; S i,e It is represented as the label attribute vector of the e-th item attribute of the i-th historically biased product among several historically biased products;

[0025] Step S206: When the similarity of the attribute details of the e-th item attribute of a certain historical biased product and a plurality of historical biased products is greater than a preset attribute details similarity threshold, the product attribute keyword group of the e-th item attribute of the certain historical biased product is recorded as the key product attribute keyword group of the e-th item attribute of the product;

[0026] Step S207: Calculate the characteristic attention value V of the e-th commodity attribute e :

[0027]

[0028] Among them, U e,i It is represented by the similarity of the attribute details of the i-th historically biased product among several historically biased products and the e-th product attribute among several historically biased products;

[0029] Step S208: When the characteristic attention degree of the e-th item commodity attribute is greater than the preset characteristic attention degree threshold, it is determined that the user has attention to the e-th item commodity attribute of the d-th item biased commodity type, and the e-th item commodity attribute is recorded as the key commodity attribute in the d-th item biased commodity type, and the key commodity attributes in several items of biased commodity types and the key commodity attribute keyword phrases of the key commodity attributes are obtained and collected to obtain the user's key commodity data;

[0030] In the above steps, the similarity of attribute details is used to calculate the user's characteristic attention to the product attributes. Because in actual practice, if the user pays attention to a certain product attribute, then when purchasing the product, the text information corresponding to a certain product attribute of the product purchased by the user will not be completely different, but will have certain characteristics. Therefore, the similarity of attribute details can be used to calculate the user's characteristic attention to the product attributes, thereby greatly increasing the accuracy of matching the product with the user in the subsequent process, and further increasing the probability of the user purchasing the product.

[0031] Furthermore, step S300 includes:

[0032] Step S301: obtaining key commodity data of a user, and obtaining preferred commodity types having key commodity attributes from the key commodity data;

[0033] Step S302: Obtain the commodities purchased by the user in each historical purchase record. When the commodity is a historically preferred commodity, record the historical purchase record as a characteristic historical purchase record;

[0034] Step S303: when the time distance between a certain historical purchase record and a characteristic historical purchase record is less than a preset unit time length, the certain historical purchase record is recorded as a marked historical purchase record of the characteristic historical purchase record;

[0035] Step S304: Analyze the collocation status between commodities of different commodity types and the degree of influence on commodity sales, wherein the collocation status between commodities of the hth commodity type and commodities of the fth commodity type and the degree of influence on commodity sales include obtaining the total number Y of characteristic historical purchase records of the hth commodity type. h, obtain each marked historical purchase record of the feature historical purchase record with the hth feature biased towards the product type, and from each marked historical purchase record, obtain the total number of marked historical purchase records with the product type of the fth product type X f ;

[0036] Step S305: Calculate the product matching value K between the product of the hth feature biased product type and the product of the fth product type h,f =X f / Y h , when the product matching value K h,f If the value is greater than the preset commodity matching threshold, the commodity with the hth feature is judged to be of the commodity type, and is matched with the commodity of the fth commodity type, which has an impact on commodity sales;

[0037] Step S306: record the fth item of product type as the matching product type of the hth item of characteristic preference product type, obtain and aggregate the matching product types of several characteristic preference product types of the user, and obtain characteristic matching product data.

[0038] Furthermore, step S400 includes:

[0039] Step S401: obtaining the user's key product data, obtaining the user's characteristic matching product data, obtaining product detail data of products sold in the product sales platform, and performing intelligent product matching for the user based on the key product data, obtaining products similar to the key product data, and recording them as recommended products;

[0040] Step S402: Acquire the user's characteristic matching data, and from the commodities sold on the commodity sales platform, obtain commodities of a commodity type that is a matching commodity type of the characteristic biased commodity type of the recommended commodity, record it as a matching recommended commodity of the recommended commodity, obtain the matching recommended commodity of the recommended commodity, and make intelligent commodity recommendations to the user on the commodity sales platform.

[0041] In order to better implement the above method, an intelligent commodity and detail identification and matching processing system based on big data is also proposed. The system includes a historical bias commodity module, a key commodity data module, a feature matching commodity data module, and an intelligent matching module;

[0042] The historical preference product module is used to evaluate the user's preference for products on the product sales platform and obtain historical preference products;

[0043] The key product data module is used to analyze the user's attention to different product attributes of product types and obtain key product data;

[0044] The feature matching product data module is used to analyze the influence of the matching status of different products on product sales and obtain the feature matching product data;

[0045] The intelligent matching module is used to match product data with preferred product data and features, intelligently match products to users, and recommend products to users on the product sales platform.

[0046] Further, the historical biased commodity module includes a type bias degree unit and a historical biased commodity unit;

[0047] A type preference degree unit is used to calculate the type preference degree of the user for each product type;

[0048] The historical preference commodity unit is used to obtain several preference commodity types of the user. When the commodity type of a commodity browsed by the user or a commodity purchased by the user is the user's preference commodity type, the commodity is recorded as a historical preference commodity.

[0049] Furthermore, the key commodity data module includes a feature attention unit and a key commodity data unit;

[0050] The characteristic attention unit is used to obtain each of the user's historical preferred products and calculate the user's characteristic attention to each product attribute within the preferred product type;

[0051] The key product data unit is used to obtain and aggregate key product attributes of several biased product types and key product attribute keyword phrases of the key product attributes to obtain the key product data of the user.

[0052] Furthermore, the characteristic matching commodity data module includes a commodity matching value unit and a characteristic matching commodity data unit;

[0053] A commodity matching value unit, used to calculate the commodity matching value between various commodity types of commodities;

[0054] The feature-matching product data unit is used to obtain and collect matching product types of several feature-biased product types of the user to obtain feature-matching product data.

[0055] Further, the intelligent matching module includes an intelligent matching unit;

[0056] The intelligent matching unit is used to intelligently match products for users based on key product data, obtain products similar to the key product data and record them as recommended products, obtain matching recommended products for the recommended products, and make intelligent product recommendations to users on the product sales platform.

[0057] Compared with the prior art, the beneficial effects of the present invention are as follows: the present invention realizes intelligent matching recommendation of user's products, analyzes the user's historical preferred products according to the user's historical browsing records and historical purchase records, obtains product detail data of the historical preferred products, analyzes the user's attention to different product attributes of product types, obtains key product attributes in several types of user's preferred products, and key product attribute keyword groups of key product attributes, and in subsequent product matching, only the text information in the detailed information of the sales product is needed to quickly find the product that the user is interested in, and also considers the influence of the matching status between products on product sales in the actual process, and by analyzing the influence of the matching status between products on product sales, the products can be recommended for matching, thereby greatly improving the sales volume of products in the product sales platform. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Figure 1 It is a method flow chart of the intelligent commodity and detail identification and matching processing system and method based on big data of the present invention;

[0059] Figure 2 It is a module diagram of the intelligent commodity and detail identification, matching and processing system and method based on big data of the present invention. DETAILED DESCRIPTION

[0060] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0061] Example: Figure 1-Figure 2 As shown, the present invention provides a technical solution, a method for intelligent commodity and detail identification and matching based on big data, the method comprising:

[0062] Step S100: Obtain the user's historical browsing records and historical purchase records, evaluate the user's preference for commodities on the commodity sales platform, and obtain historically preferred commodities;

[0063] Wherein, step S100 includes:

[0064] Step S101: Acquire each historical browsing record and each historical purchase record of the user;

[0065] Step S102: Obtain the total number α of each historical browsing record, obtain the total number β of each commodity purchase record, and calculate the characteristic coefficient γ=β / α of the historical browsing record;

[0066] Step S103: Extract the types of products browsed and purchased by the user from the historical browsing records and historical purchase records, and calculate the user's type preference for each product type, where the user's type preference for the a-th product type Q a =(C a ×γ+B a ) / (2×β), C a Indicates that the product type browsed by the user is the total number of historical browsing records of the a-th product type, B a Indicates the total number of historical purchase records of the a-th product type purchased by the user;

[0067] For example, the product type browsed by the user is, the total number of historical browsing records of the first product type C1 is 300, the product type purchased by the user is, the total number of historical purchase records of the first product type B1 is 20, the total number of purchase records of each product β is 200, and the characteristic coefficient γ of the historical browsing record is 0.25; calculate the user's type preference for the first product type Q1 = (300 × 0.25 + 20) / (2 × 200) = 0.2375;

[0068] For example, the commodity types of products include,electronic products, cosmetics, etc.;

[0069] Step S104: When the type bias degree Q a When the value is greater than the preset mark type bias degree threshold, it is determined that the user has a bias towards the a-th item product type, and the a-th item product type is recorded as the user's biased product type;

[0070] Step S105: obtaining several preferred commodity types of the user, and when the commodity type of a commodity browsed by the user or a commodity purchased by the user is the user's preferred commodity type, recording the commodity as a historical preferred commodity;

[0071] Step S200: obtaining the user's historical preference products, obtaining product detail data of the historical preference products, obtaining product attribute data from the product detail data, analyzing the user's attention to different product attributes of the product type, and obtaining key product data;

[0072] Wherein, step S200 includes:

[0073] Step S201: Acquire each of the user's historical preferred products, acquire product detail data of the historical preferred products, and acquire product attribute data from the product detail data, wherein the product attribute data includes text information corresponding to several product attributes of the preferred product type of the historical preferred products;

[0074] For example, several product attributes include color, function, and place of production;

[0075] Step S202: analyzing the user's attention to several product attributes in several preferred product types, wherein the specific analysis process of the user's attention to the e-th product attribute of the d-th preferred product type includes obtaining text information of the e-th product attribute in the historical preferred product of the d-th preferred product type, performing text preprocessing on the text information, and obtaining a product attribute keyword group;

[0076] Step S203: Obtain the total number N of several historical biased commodities of the commodity type of the dth biased commodity type d , calculate the characteristic attribute vector of each product attribute keyword in the product attribute keyword group of the e-th product attribute in the historical biased product, where the product type is, the characteristic attribute vector T of the λth product attribute keyword in the product attribute keyword group of the e-th product attribute in a historical biased product of the d-th biased product type λ :

[0077]

[0078] Among them, μ λ It is represented by the total number of occurrences of the λth product attribute keyword in the text information of the eth product attribute in a certain historically biased product; L λ It is represented by the total number of product attribute keywords in the product attribute keyword group of the e-th product attribute in a certain historically biased product; N d,λ It is represented as the total number of historically preferred products containing the λth product attribute keyword in the product attribute keyword group of the eth product attribute;

[0079] Step S204: obtaining a marked attribute vector of the e-th item of product attribute in the historically biased product based on the characteristic attribute vectors of each product attribute keyword in the product attribute keyword group of the e-th item of product attribute in the historically biased product;

[0080] Step S205: Based on the marked attribute vector, calculate the attribute details similarity U of the e-th item attribute between a certain historical biased product and several historical biased products e :

[0081]

[0082] Among them, S e It is represented as the label attribute vector of the e-th commodity attribute in a certain historical bias commodity; S i,e It is represented as the label attribute vector of the e-th item attribute of the i-th historically biased product among several historically biased products;

[0083] Step S206: When the similarity of the attribute details of the e-th item attribute of a certain historical biased product and a plurality of historical biased products is greater than a preset attribute details similarity threshold, the product attribute keyword group of the e-th item attribute of the certain historical biased product is recorded as the key product attribute keyword group of the e-th item attribute of the product;

[0084] Step S207: Calculate the characteristic attention value V of the e-th commodity attribute e :

[0085]

[0086] Among them, U e,i It is represented by the similarity of the attribute details of the i-th historically biased product among several historically biased products and the e-th product attribute among several historically biased products;

[0087] For example, U 2,1 is 0.86; U 2,2 is 0.87; U 2,3 is 0.84; N d =3; calculate the characteristic attention of the second product attribute

[0088] Step S208: When the characteristic attention degree of the e-th item commodity attribute is greater than the preset characteristic attention degree threshold, it is determined that the user has attention to the e-th item commodity attribute of the d-th item biased commodity type, and the e-th item commodity attribute is recorded as the key commodity attribute in the d-th item biased commodity type, and the key commodity attributes in several items of biased commodity types and the key commodity attribute keyword phrases of the key commodity attributes are obtained and collected to obtain the user's key commodity data;

[0089] Step S300: Obtain the user's historical purchase records, obtain historical preferred products, analyze the impact of the matching status of different products on product sales, and obtain characteristic matching product data;

[0090] Wherein, step S300 includes:

[0091] Step S301: obtaining key commodity data of a user, and obtaining preferred commodity types having key commodity attributes from the key commodity data;

[0092] Step S302: Obtain the commodities purchased by the user in each historical purchase record. When the commodity is a historically preferred commodity, record the historical purchase record as a characteristic historical purchase record.

[0093] Step S303: when the time distance between a certain historical purchase record and a characteristic historical purchase record is less than a preset unit time length, the certain historical purchase record is recorded as a marked historical purchase record of the characteristic historical purchase record;

[0094] Step S304: Analyze the collocation status between commodities of different commodity types and the degree of influence on commodity sales, wherein the collocation status between commodities of the hth commodity type and commodities of the fth commodity type and the degree of influence on commodity sales include obtaining the total number Y of characteristic historical purchase records of the hth commodity type. h , obtain each marked historical purchase record of the feature historical purchase record with the hth feature biased towards the product type, and from each marked historical purchase record, obtain the total number of marked historical purchase records with the product type of the fth product type X f ;

[0095] Step S305: Calculate the product matching value K between the product of the hth feature biased product type and the product of the fth product type h,f =X f / Y h , when the product matching value K h,f If the value is greater than the preset commodity matching threshold, the commodity with the hth feature is judged to be of the commodity type, and is matched with the commodity of the fth commodity type, which has an impact on commodity sales;

[0096] Step S306: record the fth item of product type as the collocation product type of the hth item of feature-biased product type, obtain and aggregate the collocation product types of several feature-biased product types of the user, and obtain feature-matching product data;

[0097] Step S400: Obtain the user's preferred product data and feature matching product data. The product sales platform performs intelligent product matching for the user based on the preferred product data and feature matching product data, and recommends products to the user on the product sales platform.

[0098] Wherein, step S400 includes:

[0099] Step S401: obtaining the user's key product data, obtaining the user's characteristic matching product data, obtaining product detail data of products sold in the product sales platform, and performing intelligent product matching for the user based on the key product data, obtaining products similar to the key product data, and recording them as recommended products;

[0100] Step S402: Acquire the characteristic matching data of the user, and from the commodities sold on the commodity sales platform, obtain commodities of a commodity type whose characteristic is biased towards the commodity type of the recommended commodity, record them as the matching recommended commodities of the recommended commodity, obtain the matching recommended commodities of the recommended commodity, and make intelligent commodity recommendations to the user on the commodity sales platform;

[0101] In order to better implement the above method, an intelligent commodity and detail identification and matching processing system based on big data is also proposed. The system includes a historical bias commodity module, a key commodity data module, a feature matching commodity data module, and an intelligent matching module;

[0102] The historical preference product module is used to evaluate the user's preference for products on the product sales platform and obtain historical preference products;

[0103] The key product data module is used to analyze the user's attention to different product attributes of product types and obtain key product data;

[0104] The feature matching product data module is used to analyze the influence of the matching status of different products on product sales and obtain the feature matching product data;

[0105] Intelligent matching module, used to match products to users based on preferred product data and feature matching product data, and recommend products to users on the product sales platform;

[0106] Among them, the historical biased commodity module includes a type bias degree unit and a historical biased commodity unit;

[0107] A type preference degree unit is used to calculate the type preference degree of the user for each product type;

[0108] The historical preference commodity unit is used to obtain several preference commodity types of the user. When the commodity type of a commodity browsed by the user or a commodity purchased by the user is the user's preference commodity type, the commodity is recorded as a historical preference commodity.

[0109] Among them, the key commodity data module includes a feature attention unit and a key commodity data unit;

[0110] The characteristic attention unit is used to obtain each of the user's historical preferred products and calculate the user's characteristic attention to each product attribute within the preferred product type;

[0111] The key commodity data unit is used to obtain and aggregate key commodity attributes in several biased commodity types and key commodity attribute keyword phrases of the key commodity attributes to obtain the key commodity data of the user;

[0112] Among them, the characteristic matching commodity data module includes a commodity matching value unit and a characteristic matching commodity data unit;

[0113] A commodity matching value unit, used to calculate the commodity matching value between various commodity types of commodities;

[0114] A feature matching product data unit is used to obtain and aggregate matching product types of a user's several feature-biased product types to obtain feature matching product data;

[0115] Wherein, the intelligent matching module includes an intelligent matching unit;

[0116] The intelligent matching unit is used to intelligently match products for users based on key product data, obtain products similar to the key product data and record them as recommended products, obtain matching recommended products for the recommended products, and make intelligent product recommendations to users on the product sales platform.

[0117] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. An intelligent commodity and detail identification and matching processing method based on big data, characterized in that: The method comprises: Step S100: Obtain the user's historical browsing records and historical purchase records, evaluate the user's preference for commodities on the commodity sales platform, and obtain historical preference commodities; Step S200: obtaining the user's historical preference products, obtaining product detail data of the historical preference products, obtaining product attribute data from the product detail data, analyzing the user's attention to different product attributes of product types, and obtaining key product data; Step S300: Obtain the user's historical purchase records, obtain the historical preferred commodities, analyze the influence of the matching status of different commodities on commodity sales, and obtain characteristic matching commodity data; Step S400: Obtain the user's preferred product data and feature-matched product data. The product sales platform performs intelligent product matching for the user based on the preferred product data and feature-matched product data, and recommends products to the user on the product sales platform.

2. The intelligent commodity and details identification and matching processing method based on big data according to claim 1 is characterized in that: The step S100 includes: Step S101: Acquire each historical browsing record and each historical purchase record of the user; Step S102: obtaining the total number α of each of the historical browsing records, obtaining the total number β of each of the commodity purchase records, and calculating the characteristic coefficient γ=β / α of the historical browsing records; Step S103: extract the commodity types of the commodities browsed and purchased by the user from the historical browsing records and the historical purchase records, and calculate the type preference degree of the user for each commodity type, wherein the type preference degree Q of the user for the a-th commodity type is a =(C a ×γ+B a ) / (2×β), C a Indicates that the product type browsed by the user is the total number of historical browsing records of the a-th product type, B a Indicates that the commodity type purchased by the user is the total number of historical purchase records of the a-th commodity type; Step S104: When the type bias degree Q a When the value is greater than a preset mark type bias degree threshold, it is determined that the user has a bias towards the a-th item of product type, and the a-th item of product type is recorded as the user's biased product type; Step S105: obtaining several preferred commodity types of the user, and when the commodity type of a commodity browsed by the user or a commodity purchased by the user is the user's preferred commodity type, recording the commodity as a historical preferred commodity.

3. The intelligent commodity and details identification and matching processing method based on big data according to claim 2 is characterized in that: The step S200 includes: Step S201: obtaining each of the user's historical preferred products, obtaining product detail data of the historical preferred products, and obtaining product attribute data from the product detail data, wherein the product attribute data includes text information corresponding to several product attributes of the preferred product type of the historical preferred products; Step S202: analyzing the user's attention to several commodity attributes in the several preferred commodity types, wherein the specific analysis process of the user's attention to the e-th commodity attribute of the d-th preferred commodity type includes obtaining text information of the e-th commodity attribute in the historical preferred commodity of the d-th preferred commodity type, performing text preprocessing on the text information to obtain a commodity attribute keyword group; Step S203: Obtain the total number N of several historical biased commodities of the commodity type of the dth biased commodity type. d , calculate the characteristic attribute vector of each product attribute keyword in the product attribute keyword group of the e-th product attribute in the historical biased product, wherein the product type is, in a certain historical biased product of the d-th biased product type, the characteristic attribute vector T of the λ-th product attribute keyword in the product attribute keyword group of the e-th product attribute λ : Among them, μ λ L is represented by the total number of occurrences of the λth commodity attribute keyword in the text information of the eth commodity attribute in the certain historically preferred commodity; λ N is represented by the total number of product attribute keywords in the product attribute keyword group of the e-th product attribute in the certain historically preferred product; d,λ The total number of historically preferred products containing the λth product attribute keyword in the product attribute keyword group of the eth product attribute; Step S204: obtaining a tag attribute vector of the e-th item of product attribute in the historically biased product based on the characteristic attribute vector of each product attribute keyword in the product attribute keyword group of the e-th item of product attribute in the historically biased product; Step S205: Based on the marked attribute vector, calculate the attribute details similarity U of the e-th item attribute between the certain historical biased product and the plurality of historical biased products. e : Among them, S e S is represented as a label attribute vector of the e-th commodity attribute in the certain historical bias commodity; i,e It is represented as a label attribute vector of the e-th item attribute of the i-th historically biased product among the several historically biased products; Step S206: When the similarity of the attribute details of the e-th item attribute of the certain historical biased product and the several historical biased products is greater than a preset attribute details similarity threshold, the product attribute keyword group of the e-th item attribute of the certain historical biased product is recorded as the key product attribute keyword group of the e-th item attribute of the product; Step S207: Calculate the characteristic attention value V of the e-th commodity attribute e : Among them, U e,i It is represented by the similarity of the attribute details of the e-th item attribute of the i-th historically biased product among the several historically biased products and the several historically biased products; Step S208: When the characteristic attention degree of the e-th item product attribute is greater than a preset characteristic attention degree threshold, it is determined that the user has attention to the e-th item product attribute of the d-th item biased product type, and the e-th item product attribute is recorded as the key product attribute in the d-th item biased product type. The key product attributes in several items of biased product types and the key product attribute keyword phrases of the key product attributes are obtained and aggregated to obtain the key product data of the user.

4. The intelligent commodity and details identification and matching processing method based on big data according to claim 3 is characterized in that: The step S300 includes: Step S301: Acquire key commodity data of the user, and acquire preferred commodity types having key commodity attributes from the key commodity data; Step S302: obtaining commodities purchased by the user in each historical purchase record, and when the commodity is a historically preferred commodity, recording the historical purchase record as a characteristic historical purchase record; Step S303: when the time distance between a certain historical purchase record and the characteristic historical purchase record is less than a preset unit time length, the certain historical purchase record is recorded as a marked historical purchase record of the characteristic historical purchase record; Step S304: Analyze the collocation status between commodities of different commodity types and the degree of influence on commodity sales, wherein the collocation status between commodities of the hth commodity type and commodities of the fth commodity type and the degree of influence on commodity sales include obtaining the total number Y of characteristic historical purchase records of the hth commodity type. h , obtain each marked historical purchase record of the characteristic historical purchase record of the hth feature biased towards the commodity type, and obtain the total number X of marked historical purchase records of the commodity type of the fth item from each marked historical purchase record f ; Step S305: Calculate the product matching value K between the product of the hth feature biased product type and the product of the fth product type h,f =X f / Y h , when the product matching value K h,f If the value is greater than a preset commodity matching threshold, it is determined that the commodity with the hth feature tends to be a commodity type, and is matched with the commodity with the fth feature type, which has an impact on the sales of the commodity; Step S306: record the f-th item of product type as the matching product type of the h-th item of feature-biased product type, obtain and aggregate matching product types of several feature-biased product types of the user, and obtain feature-matching product data.

5. The intelligent commodity and details identification and matching processing method based on big data according to claim 4 is characterized in that: The step S400 includes: Step S401: obtaining the key product data of the user, obtaining the characteristic matching product data of the user, obtaining the product detail data of the products sold in the product sales platform, and performing intelligent matching of products for the user based on the key product data, obtaining products similar to the key product data, and recording them as recommended products; Step S402: Acquire the characteristic matching data of the user, and from the commodities sold on the commodity sales platform, obtain commodities of a commodity type that is a matching commodity type of the characteristic-biased commodity type of the recommended commodity, record them as matching recommended commodities of the recommended commodity, obtain matching recommended commodities of the recommended commodity, and make intelligent commodity recommendations to the user on the commodity sales platform.

6. A system for intelligent commodity and details identification and matching based on big data, used to execute the intelligent commodity and details identification and matching method based on big data according to any one of claims 1 to 5, characterized in that: The system includes a historically biased commodity module, a key commodity data module, a feature matching commodity data module, and an intelligent matching module; The historical preference commodity module is used to evaluate the user's preference for commodities on the commodity sales platform and obtain historical preference commodities; The key commodity data module is used to analyze the user's attention to different commodity attributes of commodity types to obtain key commodity data; The characteristic matching commodity data module is used to analyze the influence of the matching status of different commodities on commodity sales and obtain characteristic matching commodity data; The intelligent matching module is used to intelligently match commodities for the user according to the biased commodity data and the feature matching commodity data, and to recommend commodities to the user on the commodity sales platform.

7. The intelligent commodity and details identification and matching processing system based on big data according to claim 6 is characterized in that: The historical biased commodity module includes a type bias degree unit and a historical biased commodity unit; The type preference unit is used to calculate the type preference degree of the user for each commodity type; The historical preferred product unit is used to obtain several preferred product types of the user. When the product type of a product browsed by the user or a product purchased by the user is the user's preferred product type, the product is recorded as a historical preferred product.

8. The intelligent commodity and details identification and matching processing system based on big data according to claim 6 is characterized in that: The key commodity data module includes a feature attention unit and a key commodity data unit; The characteristic attention unit is used to obtain each of the user's historical preferred commodities and calculate the characteristic attention of the user to each commodity attribute within the preferred commodity type; The key commodity data unit is used to acquire and collect key commodity attributes in a plurality of biased commodity types and key commodity attribute keyword phrases of the key commodity attributes to obtain the key commodity data of the user.

9. The intelligent commodity and details identification and matching processing system based on big data according to claim 6 is characterized in that: The characteristic matching commodity data module includes a commodity matching value unit and a characteristic matching commodity data unit; The commodity matching value unit is used to calculate the commodity matching value between various commodity types of commodities; The feature matching product data unit is used to obtain and collect matching product types of several feature-biased product types of the user, Get feature matching product data.

10. The intelligent commodity and details identification and matching processing system based on big data according to claim 6, characterized in that: The intelligent matching module includes an intelligent matching unit; The intelligent matching unit is used to intelligently match products for the user based on the key product data, obtain products similar to the key product data and record them as recommended products, obtain matching recommended products for the recommended products, and make intelligent product recommendations to the user on the product sales platform.

Images