CN116401468B - Intelligent recommendation system and method based on product concept design characteristic combination - Google Patents
Intelligent recommendation system and method based on product concept design characteristic combination Download PDFInfo
- Publication number
- CN116401468B CN116401468B CN202310405222.0A CN202310405222A CN116401468B CN 116401468 B CN116401468 B CN 116401468B CN 202310405222 A CN202310405222 A CN 202310405222A CN 116401468 B CN116401468 B CN 116401468B
- Authority
- CN
- China
- Prior art keywords
- product
- commodity
- module
- target
- information
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000013461 design Methods 0.000 title claims abstract description 41
- 238000000034 method Methods 0.000 title claims abstract description 41
- 230000004927 fusion Effects 0.000 claims abstract description 27
- 230000010365 information processing Effects 0.000 claims abstract description 22
- 230000002452 interceptive effect Effects 0.000 claims abstract description 12
- 230000006870 function Effects 0.000 claims description 37
- 238000005516 engineering process Methods 0.000 claims description 33
- 238000001914 filtration Methods 0.000 claims description 30
- 238000003058 natural language processing Methods 0.000 claims description 19
- 230000008569 process Effects 0.000 claims description 11
- 238000004458 analytical method Methods 0.000 claims description 9
- 238000002372 labelling Methods 0.000 claims description 9
- 230000001502 supplementing effect Effects 0.000 claims description 9
- 238000012545 processing Methods 0.000 claims description 7
- 238000004364 calculation method Methods 0.000 claims description 6
- 230000008451 emotion Effects 0.000 claims description 6
- 238000007726 management method Methods 0.000 claims description 6
- 238000013523 data management Methods 0.000 claims description 5
- 238000007781 pre-processing Methods 0.000 claims description 3
- 238000010845 search algorithm Methods 0.000 claims description 3
- 230000009469 supplementation Effects 0.000 claims description 3
- 238000012549 training Methods 0.000 claims description 3
- 238000010586 diagram Methods 0.000 description 5
- 239000013598 vector Substances 0.000 description 5
- 239000011159 matrix material Substances 0.000 description 3
- 238000012827 research and development Methods 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 230000008439 repair process Effects 0.000 description 2
- 238000012356 Product development Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000002860 competitive effect Effects 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000000354 decomposition reaction Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000011049 filling Methods 0.000 description 1
- 238000000227 grinding Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9536—Search customisation based on social or collaborative filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/30—Semantic analysis
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02P—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN THE PRODUCTION OR PROCESSING OF GOODS
- Y02P90/00—Enabling technologies with a potential contribution to greenhouse gas [GHG] emissions mitigation
- Y02P90/30—Computing systems specially adapted for manufacturing
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Databases & Information Systems (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Audiology, Speech & Language Pathology (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The application discloses intelligent recommendation system and method based on product concept design characteristic combination, wherein the system comprises: the system comprises an external commodity information processing module, an enterprise internal product information module, a data fusion treatment module, a conceptual design generation module and a product characteristic recommendation module; the external commodity information processing module acquires information of related commodities on the market based on external data; the enterprise internal product information acquisition module is used for acquiring the function, characteristic, technical specification and parameter information of the internal product; the data fusion treatment module performs unified standardized coding on the function, the characteristics, the technical specification and the parameter information of the internal product and the information of the external related commodity to form structured data; the user of the conceptual design generating module inputs query conditions based on the interactive operation and the structured data, and determines key characteristics of target commodities to be analyzed and target commodities; the product characteristic recommendation module recommends optimal product characteristic combinations for the user based on the target commodity key characteristics and the structured data.
Description
Technical Field
The application belongs to the technical fields of enterprise intelligent innovation, technical innovation management and electronic commerce, and particularly relates to an intelligent recommendation system and method based on product concept design characteristic combination.
Background
At present, on the aspect of new product research and development of enterprises, most of domestic enterprises still have the phenomenon of minor repair and minor repair of the existing products or follow-up imitation of competitive products based on experience, and the phenomenon directly leads to the behaviors of price reduction competition, improper marketing and the like of the domestic market in various fields, and finally influences the competitiveness of manufactured goods in China and the economic benefits of related enterprises.
At the microscopic level, the aforementioned techniques have two major problems in application: first, scene limitations: the application scene of the technology is concentrated on product recommendation of an e-commerce platform and content recommendation of each large application platform, and is rarely applied to the enterprise research and development process; this also indirectly results in inaccurate decision bases in the development decision process of enterprises, which still mainly include traditional market research results, and such paper questionnaires generally cannot accurately reflect consumer preferences for products, thereby causing directional errors in enterprise product development decisions. Second, information granularity limitations: when the technology is applied, the granularity of the processed information is coarse, and the recommended information mostly stays on the commodity or service name level and cannot be processed for the finer commodity or service information, so that the accuracy and usability of the recommended information are insufficient.
Disclosure of Invention
The application aims to solve the defects of the prior art and provides an intelligent recommendation system based on product concept design characteristic combination, which comprises:
the system comprises an external commodity information processing module, an enterprise internal product information module, a data fusion treatment module, a conceptual design generation module and a product characteristic recommendation module;
the external commodity information processing module is used for acquiring information of related commodities on the market based on external data;
the enterprise internal product information acquisition module is used for acquiring the function, characteristic, technical specification and parameter information of the internal product;
the data fusion treatment module is used for carrying out unified standardized coding on the function, the characteristics, the technical specification and the parameter information of the internal product and the information of the external related commodity to form structured data;
the user of the conceptual design generating module is used for inputting query conditions based on interactive operation and the structured data, and determining key characteristics of target commodities to be analyzed and target commodities;
the product characteristic recommending module is used for recommending optimal product characteristic combination for the user based on the key characteristics of the target commodity and the structured data.
Optionally, the external commodity information processing module comprises a similar commodity positioning sub-module, a matching and supplementing sub-module and an external data obtaining sub-module;
the similar commodity positioning sub-module is used for positioning commodities similar to a user target product in the market;
the matching supplementing submodule is used for supplementing the similar commodities by combining a fuzzy matching method;
the external data acquisition submodule is used for acquiring external commodity related data according to the similar commodity after supplementation.
Optionally, the similar commodity positioning sub-module locates similar commodities by using a clustering algorithm, and specifically includes:
manually marking key functional characteristics of the target commodity;
manually labeling the approximate corpus of the key functional characteristics;
establishing a language model of key functional characteristics of the target commodity based on a natural language processing technology and the approximate corpus;
and clustering the commodities with similar key functional characteristics of the target commodity based on a language model of the key functional characteristics of the target commodity by using a semi-supervised natural language processing technology.
Optionally, the data fusion management module comprises a picture data management sub-module, a semantic analysis sub-module and a fusion sub-module;
the picture data management sub-module is used for carrying out scene recognition on pictures, abstracting out scene subjects, matching with product functions and obtaining picture processing information based on matching information;
the semantic fusion submodule is used for carrying out theme and emotion analysis on commodity comment information;
the fusion sub-module is used for fusing the picture processing information with the text information.
Optionally, the concept design generating module supports keyword fuzzy matching based on a natural language processing method, and finally outputs a data structure of target commodities and characteristic parameters thereof which are required to be subjected to concept design by a user.
Optionally, the product characteristic recommendation module comprises a condition acquisition sub-module, a collaborative recommendation calculation sub-module and a product characteristic combination output sub-module;
the condition acquisition submodule is used for reading the data structure of the conceptual design generating module;
the collaborative recommendation calculation sub-module is used for implementing an improved collaborative filtering recommendation algorithm on the target product characteristic combination based on the data structure;
and the product characteristic combination output sub-module is used for obtaining relevant combinations and recommending according to the result of the improved collaborative filtering recommendation algorithm.
Optionally, the improved collaborative filtering recommendation algorithm specifically includes:
characteristic clustering is carried out based on the characteristic combination of the target product input by the user and the characteristic combination of the key target bid product;
filtering potential similar items on the market based on the characteristic clusters;
and carrying out association rule filtering on the similar commodities based on the transverse association rule and the longitudinal association rule to obtain a characteristic combination set.
Optionally, the transverse association rule is a multidimensional weight of product characteristics, including value, cost and technical criticality; the longitudinal association rules are multi-dimensional weights of product characteristics, including sales, concerns, and praise.
The application also includes an intelligent recommendation method based on product concept design characteristic combination, which comprises the following steps:
acquiring information of related commodities on the market based on external data;
acquiring the function, characteristic, technical specification and parameter information of an internal product;
the method comprises the steps of carrying out unified standardized coding on function, characteristics, technical specifications and parameter information of an internal product and information of an external related commodity to form structured data;
inputting query conditions based on the interactive operation and the structured data, and determining key characteristics of target commodities to be analyzed and the target commodities;
and recommending optimal product characteristic combinations for the user based on the key characteristics of the target commodity and the structured data.
Compared with the prior art, the beneficial effects of this application are:
the technical scheme is based on an advanced big data hardware system, and through careful decomposition of functions and characteristics of existing products on the market, historical development records, preferences and product technical details of enterprise users are effectively organized, and further the data are deeply analyzed by using a mainstream big data analysis technology. Meanwhile, the technical scheme improves a coordination filtering algorithm aiming at the big data, and realizes a joint decision recommendation mechanism of the technical characteristics of the multidimensional product through a self-grinding product technical characteristic clustering method and a transverse and longitudinal association rule filtering algorithm, so that the accuracy and precision of research and development recommendation of the enterprise product are improved.
Drawings
For a clearer description of the technical solutions of the present application, the drawings that are required to be used in the embodiments are briefly described below, it being evident that the drawings in the following description are only some embodiments of the present application, and that other drawings can be obtained according to these drawings without inventive effort for a person skilled in the art.
FIG. 1 is a system overall architecture diagram of an intelligent recommendation system and method based on product concept design feature combinations in an embodiment of the present application;
FIG. 2 is a schematic diagram of an external merchandise information processing module of an intelligent recommendation system and method based on combination of product concept design characteristics according to an embodiment of the present application;
FIG. 3 is a block diagram of an enterprise internal product information module and a data fusion management module of an intelligent recommendation system and method based on product concept design feature combinations according to an embodiment of the present application;
FIG. 4 is a schematic diagram of a conceptual design generating module of an intelligent recommendation system and method based on a combination of product conceptual design characteristics according to an embodiment of the present application;
fig. 5 is a schematic diagram of a product characteristic recommendation module of an intelligent recommendation system and method based on product concept design characteristic combination according to an embodiment of the present application.
Detailed Description
The following description of the embodiments of the present application will be made clearly and fully with reference to the accompanying drawings, in which it is evident that the embodiments described are only some, but not all, of the embodiments of the present application. All other embodiments, which can be made by one of ordinary skill in the art without undue burden from the present disclosure, are within the scope of the present disclosure.
In order that the above-recited objects, features and advantages of the present application will become more readily apparent, a more particular description of the invention briefly described above will be rendered by reference to specific embodiments that are illustrated in the appended drawings.
Example 1
In this embodiment, as shown in fig. 1, an intelligent recommendation system and method based on product concept design feature combination specifically includes:
the system comprises an external commodity information processing module, an enterprise internal product information module, a data fusion treatment module, a conceptual design generation module and a product characteristic recommendation module;
as shown in fig. 2, the external commodity information processing module is used for acquiring information of related commodities on the market based on external data; optionally, the external commodity information processing module comprises a similar commodity positioning sub-module, a matching and supplementing sub-module and an external data obtaining sub-module; the similar commodity positioning sub-module is used for positioning commodities similar to a user target product in the market; the matching and supplementing submodule is used for supplementing the similar commodities by combining a fuzzy matching method; the external data acquisition submodule is used for acquiring external commodity related data according to the similar commodity after supplementation.
Specifically, the similar commodity positioning sub-module uses a clustering algorithm to position similar commodities, including: manually marking key functional characteristics of the target commodity; manually labeling the approximate corpus of the key functional characteristics; establishing a language model of key functional characteristics of the target commodity based on a natural language processing technology and the approximate corpus; and clustering the commodities with similar key functional characteristics of the target commodity based on a language model of the key functional characteristics of the target commodity by using a semi-supervised natural language processing technology.
As shown in fig. 3, the intra-enterprise product information obtaining module is configured to obtain function, characteristic, technical specification and parameter information of an internal product;
the data fusion management module is used for carrying out unified standardized coding on the function, the characteristics, the technical specification and the parameter information of the internal product and the information of the external related commodity to form structured data. And carrying out data fusion treatment on the external commodity data input by the external commodity information processing module and the internal product information obtained by the internal product information obtaining module of the enterprise, forming standardized product characteristics, technology, sales volume and public praise indexes for each commodity, and outputting the fused data to the product characteristic recommending module.
Wherein the unified standardized coding process comprises: the product, technology and knowledge categories of a target enterprise are manually researched, and based on a product-function-technology-knowledge system architecture, a knowledge map of the enterprise product is established and encoded by combining manual expert labeling with natural language processing semi-supervised learning technology.
For example, a product has 20 functions, then we need 2 digits to encode the functions, then a function is designed to 9 technologies, then we need 1 digit encoding, then one technology involves 120 related key knowledge, then we also need 3 digits encoding, then the product can be represented by 6 digits encoding, and the encoding can locate not only the product, but also all functions, technologies and knowledge related to the product.
The method completely and accurately decomposes the enterprise products, changes the enterprise from products to functions to technologies to just complete structuring, and simultaneously, the encoding is more convenient for inquiring big data.
Specifically, the fusion treatment is mainly aimed at word and picture data treatment and semantic fusion, and comprises the following steps:
1. manually marking key functional characteristics of the target commodity;
2. manually labeling the approximate corpus of the key functional characteristics generated in the previous step;
3. establishing a language model of key functional characteristics of the commodity based on a natural language processing technology;
4. clustering commodities with similar key functional characteristics of target commodities based on a semi-supervised natural language processing technology;
5. keyword matching is carried out on the commodity information obtained through searching, and the price and sales of the product are obtained;
6. carrying out theme and emotion analysis on the commodity comment information obtained by searching;
7. keyword matching is carried out on technical details in the commodity information obtained through searching;
8. processing the picture information;
9. carrying out scene recognition on the picture, abstracting out a scene main body, and matching with the product function;
10. OCR recognition is carried out on the embedded characters of the picture, and the character is matched with the functions and the technology of the product;
11. and fusing the picture identification result with the character information identification result.
As shown in fig. 4, the user of the conceptual design generating module is configured to determine the key characteristics of the target commodity to be analyzed and the target commodity based on the interactive operation and the query condition input by the structured data; the concept design generating module supports keyword fuzzy matching based on a natural language processing method, and finally, a data structure of target commodities and characteristic parameters thereof which are required to be subjected to concept design by a user is output.
As shown in fig. 5, the product characteristic recommendation module is configured to recommend an optimal product characteristic combination for the user based on the key characteristics of the target commodity and the structured data.
Specifically, the product characteristic recommendation module comprises a condition acquisition sub-module, a collaborative recommendation calculation sub-module and a product characteristic combination output sub-module; the condition acquisition submodule is used for reading the data structure of the conceptual design generating module; the collaborative recommendation calculation sub-module is used for implementing an improved collaborative filtering recommendation algorithm on the target product characteristic combination based on the data structure; and the product characteristic combination output sub-module is used for obtaining relevant combinations and recommending according to the result of the improved collaborative filtering recommendation algorithm.
The improved collaborative filtering recommendation algorithm specifically comprises the following steps: characteristic clustering is carried out based on the characteristic combination of the target product input by the user and the characteristic combination of the key target bid product; filtering potential similar items on the market based on the characteristic clusters; and carrying out association rule filtering on the similar commodities based on the transverse association rule and the longitudinal association rule to obtain a characteristic combination set.
Specifically, the transverse association rule is a multidimensional weight of product characteristics, including value, cost and technical key degree; the longitudinal association rules are multi-dimensional weights of product characteristics, including sales, concerns, and praise.
Example two
The application also includes an intelligent recommendation method based on product concept design characteristic combination, the method includes the following steps:
acquiring information of related commodities on the market based on external data; specifically, an improved product property search algorithm is used, and the core innovative logic of the algorithm is: and (3) positioning commodities similar to the user target products in the market on the product function and characteristic level by utilizing a clustering algorithm in an external data source, and adding a keyword fuzzy matching method to improve the searching accuracy. And further obtains data of the commodity including price, sales volume, public praise, functions, technical parameters and the like, and inputs the external data to the internal product technical information processing module.
The algorithm specifically comprises the following steps:
1. manually marking key functional characteristics of the target commodity;
2. manually labeling the approximate corpus of the key functional characteristics generated in the previous step;
3. establishing a language model of key functional characteristics of the commodity based on a natural language processing technology;
4. clustering commodities similar to key functional characteristics of target commodities based on a semi-supervised natural language processing technology;
5. keyword matching is carried out on the commodity information obtained through searching, and the price and sales of the product are obtained;
6. carrying out theme and emotion analysis on the commodity comment information obtained by searching;
7. and carrying out keyword matching on technical details in the commodity information obtained by searching.
Specifically, the keyword fuzzy matching process includes: 1. determining a target keyword set according to the characteristics, the functions and the like of the commodity, and preprocessing a corpus formed by the captured commodity network propaganda text content;
2. performing similarity matching on each target keyword in a corpus based on indexes such as editing distance, cosine similarity, jaccard similarity and the like, and calculating similarity scores of the target keywords;
3. based on the similarity score, training a fuzzy degree threshold value, and adjusting and filtering the number of approximate words of the target keywords in the corpus to obtain words with high similarity with the target keywords.
The method comprises the steps of carrying out unified standardized coding on function, characteristics, technical specifications and parameter information of an internal product and information of an external related commodity to form structured data; acquiring the function, characteristic, technical specification and parameter information of an internal product; specifically, a unified standardized coding system for functions, characteristics, technical specifications and parameters of all products in an enterprise is constructed first. On the basis, the external commodity data input by the external commodity information processing module and the internal product technical data input by the external commodity information processing module are processed, data fusion treatment is carried out, standardized product characteristics, technology, sales volume and public praise indexes are formed for each commodity, and the fused data are output to the product characteristic recommending module.
Inputting query conditions based on the interactive operation and the structured data, and determining key characteristics of target commodities to be analyzed and the target commodities; the user inputs inquiry conditions for commodities through the interactive interface, and the inquiry conditions are uniformly standardized and encoded according to functions, characteristics, technical specifications and parameters of all products in the enterprise generated by the internal product technical information processing module
And recommending optimal product characteristic combinations for the user based on the key characteristics of the target commodity and the structured data. Reading the output of a conceptual design generating module, and implementing an improved collaborative filtering recommendation algorithm on the characteristic combination of a target product, wherein the core innovation logic of the algorithm is as follows: the traditional recommendation algorithm recommends commodity contents to users, scores the favorites of the commodity contents representing the users, and recommends based on the similarity of the scores. The improved recommendation algorithm provided by the patent adopts a collaborative filtering algorithm of improved clustering combined with association rules, and aims to improve the accuracy of recommendation. Firstly, characteristic clustering is carried out based on characteristic combinations of target products input by a user and characteristic combinations of key target bid products, and the specific characteristic clustering process is as follows: representing the functional characteristics of the target product and the on-sale commodity as characteristic vectors based on the product structural codes, wherein each dimension represents that the commodity is a certain functional characteristic; calculating the similarity between the characteristic vectors, taking the product characteristic vectors as input data, and calculating the distance between the characteristic vectors by using the text editing distance based on the coding difference; hierarchical clustering is carried out based on the characteristic vector distance, the semi-supervised learning technology is utilized to adjust the parameters of the hierarchical clustering algorithm, and the clustering result is optimized; and finding a commodity set which is closer to the characteristic combination of the target product based on the clustering result.
Filtering the potential similar goods in the market. Then, based on the transverse association rule (the dimensions of multi-dimensional weight, value, cost, technical key degree and the like of the product characteristics) and the longitudinal association rule (the dimensions of multi-dimensional weight, sales, attention, public praise and the like of the product characteristics), carrying out association rule filtering, wherein the specific process of association rule filtering comprises the following steps: 1. obtaining commodities on the market which are relatively close to the target characteristic combination of the user through the characteristic clustering; 2. based on the external commodity information processing module, two attribute matrices for all such commodities can be obtained: a transverse attribute matrix, which acts on each such commodity, and is listed as a value, cost and technical criticality attribute, wherein the attribute value comes from the internal technical score of an enterprise in the structured coding system; 3. a longitudinal attribute matrix, wherein each commodity is listed as sales, attention and public praise attributes, and attribute values of the longitudinal attribute matrix come from an external commodity information processing module; 4. based on the result of the step 1, filling predictive attribute values for the characteristic set input by the enterprise user by utilizing the transverse and longitudinal attribute matrixes; 5. based on the transverse and longitudinal attribute moments, re-combining the user target characteristics and similarity of commodities on a relatively close market; 6. and re-clustering the commodities based on the similarity in the last step, and finding out commodities and characteristic combination sets thereof close to the user target characteristic combination.
And obtaining the potential optimal commodity and the characteristic combination set thereof. Finally, through product target constraint data (constraint of cost, key characteristics, technical parameters and the like) input by a user, the target constraint data is that a group of parameters are input by an enterprise user when the system is used for displaying the preference of the user to the product, and the target constraint data comprises the following components: upper cost limit, target sales, key product property 1, product property 1 core technical parameters, … …, key product property n, product property n core technical parameters.
And (3) completing the search of the optimal product characteristic combination in the potential optimal commodity and the characteristic combination set thereof, and recommending relevant combinations. Specifically, the combination set in this embodiment is: each feature set is a set of product features, each feature set may essentially constitute a piece of product from which enterprise users may develop product prototypes.
In this embodiment, the relevant combinations are a plurality of recommended combinations of the final output product functional characteristics of the system and a recommended sequence thereof, the number of the output combinations depends on parameter thresholds filtered by the clustering and association rules, each characteristic combination can basically form a product, and enterprise users can develop product prototype design according to the characteristic combinations.
The foregoing embodiments are merely illustrative of the preferred embodiments of the present application and are not intended to limit the scope of the present application, and various modifications and improvements made by those skilled in the art to the technical solutions of the present application should fall within the protection scope defined by the claims of the present application.
Claims (5)
1. An intelligent recommendation system based on product concept design characteristic combination, which is characterized by comprising: the system comprises an external commodity information processing module, an enterprise internal product information module, a data fusion treatment module, a conceptual design generation module and a product characteristic recommendation module;
the external commodity information processing module is used for acquiring information of related commodities on the market based on external data;
obtaining relevant commodity information by using an improved product characteristic searching algorithm; the improved product property search algorithm includes: manually marking key functional characteristics of the target commodity; manually labeling the approximate corpus of the key functional characteristics generated in the previous step; establishing a language model of key functional characteristics of the commodity based on a natural language processing technology; clustering commodities similar to key functional characteristics of target commodities based on a semi-supervised natural language processing technology; keyword matching is carried out on the commodity information obtained through searching, and the price and sales of the product are obtained; carrying out theme and emotion analysis on the commodity comment information obtained by searching; keyword matching is carried out on technical details in the commodity information obtained through searching; the target commodity is a commodity similar to a user target product in the market in the aspects of product functions and characteristics by utilizing a clustering algorithm in an external data source;
the enterprise internal product information acquisition module is used for acquiring the function, characteristic, technical specification and parameter information of the internal product;
the data fusion treatment module is used for carrying out unified standardized coding on the function, the characteristics, the technical specification and the parameter information of the internal product and the information of the external related commodity to form structured data;
the working process of the data fusion treatment module comprises the following steps: carrying out data fusion treatment on the external commodity data input by the external commodity information processing module and the internal product information obtained by the internal product information obtaining module of the enterprise, forming standardized product characteristics, technology, sales volume and public praise indexes for each commodity, and outputting the fused data to a product characteristic recommending module;
inputting query conditions based on the interactive operation and the structured data, and determining key characteristics of target commodities to be analyzed and the target commodities; the method comprises the steps that a user inputs inquiry conditions for commodities through an interactive interface, the inquiry conditions carry out unified standardized coding on functions, characteristics, technical specifications and parameters of products, and the products are generated by an internal product technical information processing module;
the unified standardized coding process comprises the following steps: manually researching the product, technology and knowledge categories of a target enterprise, based on a system architecture of product-function-technology-knowledge, establishing a knowledge map of the enterprise product by combining manual expert labeling with natural language processing semi-supervised learning technology, and coding; the fusion management aims at word and picture data management and semantic fusion, and comprises the following steps:
processing the picture information;
carrying out scene recognition on the picture, abstracting out a scene main body, and matching with the product function;
OCR recognition is carried out on the embedded characters of the picture, and the character is matched with the functions and the technology of the product;
fusing the picture identification result with the character information identification result; the user of the conceptual design generating module is used for inputting query conditions based on interactive operation and the structured data, and determining key characteristics of target commodities to be analyzed and target commodities;
the concept design generation module supports keyword fuzzy matching based on a natural language processing method, and finally outputs a data structure of target commodities and characteristic parameters thereof which are required to be subjected to concept design by a user;
the process of fuzzy matching of keywords comprises the following steps: determining a target keyword set according to the characteristics and the function description of the commodity, and preprocessing a corpus formed by the captured commodity network propaganda text content;
based on the editing distance, cosine similarity and Jaccard similarity index, performing similarity matching on each target keyword in a corpus, and calculating similarity scores of the target keywords;
based on the similarity score, training a fuzzy degree threshold value, and adjusting and filtering the number of approximate words of the target keywords in the corpus to obtain words with high similarity with the target keywords;
the product characteristic recommending module is used for recommending optimal product characteristic combinations for users based on the key characteristics and the structured data of the target commodity;
the product characteristic recommendation module comprises a condition acquisition sub-module, a collaborative recommendation calculation sub-module and a product characteristic combination output sub-module;
the condition acquisition submodule is used for reading the data structure of the conceptual design generating module;
the collaborative recommendation calculation sub-module is used for implementing an improved collaborative filtering recommendation algorithm on the target product characteristic combination based on the data structure;
the product characteristic combination output submodule is used for obtaining relevant combinations and recommending according to the result of the improved collaborative filtering recommendation algorithm;
the improved collaborative filtering recommendation algorithm specifically comprises the following steps: characteristic clustering is carried out based on the characteristic combination of the target product input by the user and the characteristic combination of the key target bid product; filtering potential similar items on the market based on the characteristic clusters; and carrying out association rule filtering on the similar commodities based on the transverse association rule and the longitudinal association rule to obtain a characteristic combination set.
2. The intelligent recommendation system based on combination of product concept design features according to claim 1, wherein the external commodity information processing module comprises a similar commodity positioning sub-module, a matching supplementing sub-module and an external data obtaining sub-module;
the similar commodity positioning sub-module is used for positioning commodities similar to a user target product in the market;
the matching supplementing submodule is used for supplementing the similar commodities by combining a fuzzy matching method;
the external data acquisition submodule is used for acquiring external commodity related data according to the similar commodity after supplementation.
3. The intelligent recommendation system based on product concept design characteristic combination according to claim 1, wherein the data fusion governance module comprises a picture data governance sub-module, a semantic analysis sub-module and a fusion sub-module;
the picture data management sub-module is used for carrying out scene recognition on pictures, abstracting out scene subjects, matching with product functions and obtaining picture processing information based on matching information;
the semantic analysis submodule is used for carrying out theme and emotion analysis on commodity comment information;
the fusion sub-module is used for fusing the picture processing information with the text information.
4. The intelligent recommendation system based on product concept design feature combinations of claim 3, wherein the lateral association rules are multi-dimensional weights of product features, including value, cost, and technical criticality; the longitudinal association rules are multi-dimensional weights of product characteristics, including sales, concerns, and praise.
5. The intelligent recommendation method based on the product concept design characteristic combination is characterized by comprising the following steps of:
acquiring information of related commodities on the market based on external data;
obtaining relevant commodity information by using an improved product characteristic searching algorithm; the improved product property search algorithm includes: manually marking key functional characteristics of the target commodity; manually labeling the approximate corpus of the key functional characteristics generated in the previous step; establishing a language model of key functional characteristics of the commodity based on a natural language processing technology; clustering commodities similar to key functional characteristics of target commodities based on a semi-supervised natural language processing technology; keyword matching is carried out on the commodity information obtained through searching, and the price and sales of the product are obtained; carrying out theme and emotion analysis on the commodity comment information obtained by searching; keyword matching is carried out on technical details in the commodity information obtained through searching; the target commodity is a commodity similar to a user target product in the market in the aspects of product functions and characteristics by utilizing a clustering algorithm in an external data source;
acquiring the function, characteristic, technical specification and parameter information of an internal product;
the method comprises the steps of carrying out unified standardized coding on function, characteristics, technical specifications and parameter information of an internal product and information of an external related commodity to form structured data;
carrying out data fusion treatment on the input external commodity data and the obtained internal product information, and forming standardized product characteristics, technology, sales volume and public praise indexes for each commodity;
inputting query conditions based on the interactive operation and the structured data, and determining key characteristics of target commodities to be analyzed and the target commodities; the method comprises the steps that a user inputs inquiry conditions for commodities through an interactive interface, the inquiry conditions carry out unified standardized coding on functions, characteristics, technical specifications and parameters of products, and the products are generated by an internal product technical information processing module;
the unified standardized coding process comprises the following steps: manually researching the product, technology and knowledge categories of a target enterprise, based on a system architecture of product-function-technology-knowledge, establishing a knowledge map of the enterprise product by combining manual expert labeling with natural language processing semi-supervised learning technology, and coding; the fusion management aims at word and picture data management and semantic fusion, and comprises the following steps:
processing the picture information;
carrying out scene recognition on the picture, abstracting out a scene main body, and matching with the product function;
OCR recognition is carried out on the embedded characters of the picture, and the character is matched with the functions and the technology of the product;
fusing the picture identification result with the character information identification result; the user of the conceptual design generating module is used for inputting query conditions based on interactive operation and the structured data, and determining key characteristics of target commodities to be analyzed and target commodities;
fuzzy matching of keywords based on a natural language processing method, and finally outputting a data structure of target commodities and characteristic parameters thereof which are required to be subjected to conceptual design by a user;
the process of fuzzy matching of keywords comprises the following steps: determining a target keyword set according to the characteristics and the function description of the commodity, and preprocessing a corpus formed by the captured commodity network propaganda text content;
based on the editing distance, cosine similarity and Jaccard similarity index, performing similarity matching on each target keyword in a corpus, and calculating similarity scores of the target keywords;
based on the similarity score, training a fuzzy degree threshold value, and adjusting and filtering the number of approximate words of the target keywords in the corpus to obtain words with high similarity with the target keywords;
recommending optimal product characteristic combinations for users based on the key characteristics and the structured data of the target commodity;
reading a data structure of the conceptual design generating module;
implementing an improved collaborative filtering recommendation algorithm for a combination of target product characteristics based on the data structure;
obtaining relevant combinations and recommending according to the result of the improved collaborative filtering recommendation algorithm;
the improved collaborative filtering recommendation algorithm specifically comprises the following steps: characteristic clustering is carried out based on the characteristic combination of the target product input by the user and the characteristic combination of the key target bid product; filtering potential similar items on the market based on the characteristic clusters; and carrying out association rule filtering on the similar commodities based on the transverse association rule and the longitudinal association rule to obtain a characteristic combination set.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310405222.0A CN116401468B (en) | 2023-04-17 | 2023-04-17 | Intelligent recommendation system and method based on product concept design characteristic combination |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310405222.0A CN116401468B (en) | 2023-04-17 | 2023-04-17 | Intelligent recommendation system and method based on product concept design characteristic combination |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116401468A CN116401468A (en) | 2023-07-07 |
| CN116401468B true CN116401468B (en) | 2023-12-19 |
Family
ID=87010242
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310405222.0A Active CN116401468B (en) | 2023-04-17 | 2023-04-17 | Intelligent recommendation system and method based on product concept design characteristic combination |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116401468B (en) |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR100687906B1 (en) * | 2005-12-31 | 2007-02-27 | 주식회사지앤지커머스 | System for recommendation the goods and method therefor |
| WO2009097581A1 (en) * | 2008-01-31 | 2009-08-06 | American International Group, Inc. | Method and system of developing a product |
| CN102930389A (en) * | 2012-10-23 | 2013-02-13 | 北京理工大学 | Product design knowledge management method and system |
| CN103995886A (en) * | 2014-05-30 | 2014-08-20 | 北京理工大学 | Multidimensional product design knowledge pushing frame and construction method |
| WO2014157762A1 (en) * | 2013-03-28 | 2014-10-02 | 삼성전자 주식회사 | Smart design system for providing mobile application design and simulation function, business model including same, and method for operating smart design system |
| CN104899242A (en) * | 2015-03-10 | 2015-09-09 | 四川大学 | Mechanical product design two-dimensional knowledge pushing method based on design intent |
| CN106126885A (en) * | 2016-06-16 | 2016-11-16 | 中国船舶工业综合技术经济研究院 | A kind of product-design knowledge method of servicing and device |
| CN109902116A (en) * | 2019-01-29 | 2019-06-18 | 合肥工业大学 | A kind of Eco-Design knowledge active pushing system and method |
| JP2019148936A (en) * | 2018-02-26 | 2019-09-05 | 株式会社日立製作所 | Product design support system |
| KR102221098B1 (en) * | 2020-09-08 | 2021-02-26 | 한국과학기술정보연구원 | Enterprise-customized recommending apparatus for new drug development, and control method thereof |
| CN114741603A (en) * | 2022-04-25 | 2022-07-12 | 大连理工大学 | Mixed collaborative filtering recommendation algorithm based on user clustering and commodity clustering |
| CN115392592A (en) * | 2022-10-25 | 2022-11-25 | 浪潮电子信息产业股份有限公司 | Storage product parameter configuration recommendation method, device, equipment and medium |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11157985B2 (en) * | 2017-11-29 | 2021-10-26 | Ditto Technologies, Inc. | Recommendation system, method and computer program product based on a user's physical features |
-
2023
- 2023-04-17 CN CN202310405222.0A patent/CN116401468B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR100687906B1 (en) * | 2005-12-31 | 2007-02-27 | 주식회사지앤지커머스 | System for recommendation the goods and method therefor |
| WO2009097581A1 (en) * | 2008-01-31 | 2009-08-06 | American International Group, Inc. | Method and system of developing a product |
| CN102930389A (en) * | 2012-10-23 | 2013-02-13 | 北京理工大学 | Product design knowledge management method and system |
| WO2014157762A1 (en) * | 2013-03-28 | 2014-10-02 | 삼성전자 주식회사 | Smart design system for providing mobile application design and simulation function, business model including same, and method for operating smart design system |
| CN103995886A (en) * | 2014-05-30 | 2014-08-20 | 北京理工大学 | Multidimensional product design knowledge pushing frame and construction method |
| CN104899242A (en) * | 2015-03-10 | 2015-09-09 | 四川大学 | Mechanical product design two-dimensional knowledge pushing method based on design intent |
| CN106126885A (en) * | 2016-06-16 | 2016-11-16 | 中国船舶工业综合技术经济研究院 | A kind of product-design knowledge method of servicing and device |
| JP2019148936A (en) * | 2018-02-26 | 2019-09-05 | 株式会社日立製作所 | Product design support system |
| CN109902116A (en) * | 2019-01-29 | 2019-06-18 | 合肥工业大学 | A kind of Eco-Design knowledge active pushing system and method |
| KR102221098B1 (en) * | 2020-09-08 | 2021-02-26 | 한국과학기술정보연구원 | Enterprise-customized recommending apparatus for new drug development, and control method thereof |
| CN114741603A (en) * | 2022-04-25 | 2022-07-12 | 大连理工大学 | Mixed collaborative filtering recommendation algorithm based on user clustering and commodity clustering |
| CN115392592A (en) * | 2022-10-25 | 2022-11-25 | 浪潮电子信息产业股份有限公司 | Storage product parameter configuration recommendation method, device, equipment and medium |
Non-Patent Citations (2)
| Title |
|---|
| FEATURE: New Product Development Using Feature-Drift Based Framework for Unique Aspect Recommendation;Monireh Alsadat Mirtalaie.et al;2016 IEEE 13th International Conference on e-Business Engineering (ICEBE);第24-31页 * |
| 面向客户满意的消费类电子产品概念设计过程若干关键技术研究;杨涛;中国博士学位论文全文数据库经济与管理科学辑;第81-96页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116401468A (en) | 2023-07-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110910243B (en) | Property right transaction method based on reconfigurable big data knowledge map technology | |
| CN109903117B (en) | Knowledge graph processing method and device for commodity recommendation | |
| CN110633373B (en) | Automobile public opinion analysis method based on knowledge graph and deep learning | |
| CN107705066B (en) | Information input method and electronic equipment during commodity warehousing | |
| CN107220365B (en) | Accurate recommendation system and method based on collaborative filtering and association rule parallel processing | |
| CN111898031B (en) | Method and device for obtaining user portrait | |
| CN113343125B (en) | Academic accurate recommendation-oriented heterogeneous scientific research information integration method and system | |
| CN108319734A (en) | A kind of product feature structure tree method for auto constructing based on linear combiner | |
| CN111242679A (en) | Sales forecasting method based on product review viewpoint mining | |
| CN111339284A (en) | Product intelligent matching method, device, equipment and readable storage medium | |
| CN112862569B (en) | Product appearance style evaluation method and system based on image and text multi-modal data | |
| CN111325571A (en) | Method, device and system for automatically generating commodity comment labels for multitask learning | |
| CN111858940A (en) | Multi-head attention-based legal case similarity calculation method and system | |
| CN114238577B (en) | Multi-task learning emotion classification method integrating multi-head attention mechanism | |
| CN111401936A (en) | Recommendation method based on comment space and user preference | |
| CN112163160A (en) | Knowledge graph-based sensitive identification method | |
| CN114626909A (en) | Commodity recommendation method, system, equipment and medium | |
| CN112906396A (en) | Cross-platform commodity matching method and system based on natural language processing | |
| CN114861050A (en) | Feature fusion recommendation method and system based on neural network | |
| CN112231593A (en) | Financial information intelligent recommendation system | |
| CN117436446A (en) | Weak supervision-based agricultural social sales service user evaluation data analysis method | |
| CN113592512A (en) | Online commodity identity uniqueness identification and confirmation system | |
| CN116401468B (en) | Intelligent recommendation system and method based on product concept design characteristic combination | |
| Rasouli et al. | Designing an agent for information extraction from persian e-shops | |
| CN116205212A (en) | Bid file information extraction method, device, equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |