Method and device for realizing product matching of business opportunity information

Abstract

The invention relates to a method and device for matching business opportunity information with a product, and belongs to the technical field of matching the business opportunity information with the product, and the method comprises the steps: obtaining the business opportunity information, carrying out the preprocessing of the business opportunity information, and constructing a feature lexicon based on the preprocessed business opportunity information; vectorizing each keyword in the feature word bank, and setting a weight for each vectorized keyword to construct a business opportunity vector; obtaining product information, and preprocessing the product information; converting the preprocessed product information into product information feature vectors, and constructing a product vector library; and calculating the similarity between the business opportunity information vector and a product information feature vector in a vector database, and screening texts or paragraphs with the similarity greater than a preset threshold value as products matched with the business opportunity information for recommendation. By calculating the similarity between the business opportunity information vector and the product information feature vector in the vector database, the product which most meets the user demand is matched, so that the product recommendation efficiency and precision are improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of business opportunity information matching products, and in particular, relates to a method and device for implementing business opportunity information matching products. Background Art

[0002] At present, there are difficulties in accurately matching business opportunity information with software capability platform products and multiple pain points in the manual offline search process: long time consumption, low efficiency, subjective judgment bias, etc. It is mainly reflected in the following points:

[0003] 1. Time-consuming and inefficient: Manual offline product retrieval requires comparing the business opportunity description with the product functions, applicable scope and other information in the mall one by one. The process is cumbersome and time-consuming. When faced with large-scale business opportunities, it is very easy to have problems such as incomplete information coverage and low efficiency.

[0004] 2. Subjective judgment bias: Different people may have different understandings of business opportunity needs and product features. This subjectivity may lead to inconsistency and uncertainty in matching results, affecting the final recommendation effect and user experience.

[0005] 3. Understanding semantic limitations: Traditional business opportunity systems mainly process text based on algorithms such as keyword matching and TF-IDF (term frequency-inverse document frequency). They rely on pre-built indexes and content in databases and lack an understanding of the deep semantics of the text. Summary of the invention

[0006] In view of the above deficiencies in the prior art, the purpose of the invention is to provide a method and device for matching business opportunity information with products, vectorize business opportunity information to construct a business opportunity vector, vectorize product information to construct a product information feature vector, build a product vector library based on the product information feature vector, calculate similarity to match the product that best meets user needs, thereby improving the efficiency and accuracy of product recommendations.

[0007] A first aspect of the present invention provides a method for matching business opportunity information with products, comprising:

[0008] Acquire business opportunity information, pre-process the business opportunity information, and build a feature word library based on the pre-processed business opportunity information;

[0009] Vectorize each keyword in the feature word library, set weights for each vectorized keyword and construct a business opportunity vector;

[0010] Obtain product information and pre-process the product information;

[0011] Convert the preprocessed product information into a product information feature vector, and build a product vector library based on the product information feature vector;

[0012] The similarity between the business opportunity information vector and the product information feature vector in the vector database is calculated, and the text or paragraphs with similarity greater than a preset threshold are filtered out as products matching the business opportunity information for recommendation.

[0013] Furthermore, in the above method for matching business opportunity information with products, obtaining business opportunity information, preprocessing the business opportunity information, and building a feature word library based on the preprocessed business opportunity information include:

[0014] Connect to the business opportunity system through the interface to collect key information and obtain business opportunity information;

[0015] Clear invalid or duplicate data records in business opportunity information;

[0016] Implement text segmentation operations, identify and extract key terms in business opportunity information, perform part-of-speech tagging on key terms, and filter the tagged key terms to obtain keywords;

[0017] Build a feature vocabulary based on keywords;

[0018] Classify the keywords in the feature word library according to the industry knowledge base to distinguish industry categories and demand types;

[0019] Among them, the key information includes at least: business opportunity title, customer needs, industry category and estimated amount.

[0020] Furthermore, in the above method for matching business opportunity information to products, each keyword in the feature word library is vectorized, and a weight is set for each vectorized keyword to construct a business opportunity vector, including:

[0021] Use a pre-trained text embedding model to convert each keyword in the feature vocabulary into a vector representation;

[0022] Set a weight for each vectorized keyword, and combine all feature vectors to obtain the business opportunity vector by weighted summation;

[0023] The weight is set based on the frequency of occurrence of keywords in the industry and their criticality in the business opportunity description.

[0024] Furthermore, in the above method for matching business opportunity information with products, obtaining product information and preprocessing the product information include:

[0025] Extract product details from the software capability mall;

[0026] Remove noise, duplicates and irrelevant information from product details;

[0027] Determine the size of the document length of the product details after removing noise, duplicates and irrelevant information and the size of the preset threshold;

[0028] If the judgment result is that the document length of the product details is greater than the preset threshold, the document is divided into multiple text blocks;

[0029] Among them, product details include at least: product name, description, industry and sales data.

[0030] Furthermore, in the above method for matching business opportunity information with products, the preprocessed product information is converted into a product information feature vector, and a product vector library is constructed based on the product information feature vector, including:

[0031] Perform word segmentation and keyword extraction on the pre-processed product information to obtain product information features;

[0032] Use the pre-trained text embedding model to convert product information features into vector representation to obtain product information feature vector;

[0033] Build a dictionary based on the product information feature vector, and map each word in the dictionary to an integer to get the dictionary index;

[0034] For each word with dictionary index i in the dictionary, obtain two sets of word vector representations when it is the center word and background word respectively;

[0035] Build a product vector library based on two sets of word vectors;

[0036] Here, i represents an integer.

[0037] Furthermore, in the above method for matching business opportunity information with products, the similarity between the business opportunity information vector and the product information feature vector in the vector database is calculated by cosine similarity.

[0038] Furthermore, in the above method for matching business opportunity information with products, the text or paragraphs with a similarity greater than a preset threshold are screened and recommended as products matching the business opportunity information, including:

[0039] The text or paragraphs with similarity greater than a preset threshold are taken as preliminary matching results;

[0040] The preliminary matching results are arranged from high to low according to the similarity score, and the text or paragraph with the highest similarity score is recommended as the product matching the business opportunity information.

[0041] The second aspect of the present invention further provides a device for matching business opportunity information with products, comprising:

[0042] Acquisition module: used to acquire business opportunity information, preprocess the business opportunity information, and build a feature word library based on the preprocessed business opportunity information;

[0043] Construction module: used to vectorize each keyword in the feature vocabulary, set weights for each vectorized keyword and construct business opportunity vectors;

[0044] Preprocessing module: used to obtain product information and preprocess the product information;

[0045] Conversion module: used to convert the preprocessed product information into product information feature vectors, and build a product vector library based on the product information feature vectors;

[0046] Matching module: used to calculate the similarity between the business opportunity information vector and the product information feature vector in the vector database, and filter the text or paragraphs with similarity greater than a preset threshold as products matching the business opportunity information for recommendation.

[0047] According to a third aspect of the present invention, an electronic device is provided, comprising: a processor and a memory;

[0048] The processor is used to execute any one of the above methods for matching business opportunity information with products by calling the program or instruction stored in the memory.

[0049] In a fourth aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a program or instruction, and the program or instruction enables a computer to execute any one of the above methods for matching business opportunity information with products.

[0050] The beneficial effects of the present invention are as follows: the present invention obtains business opportunity information, pre-processes the business opportunity information, and constructs a feature word library based on the pre-processed business opportunity information; vectorizes each keyword in the feature word library, sets weights for each vectorized keyword to construct a business opportunity vector; obtains product information, and pre-processes the product information; converts the pre-processed product information into a product information feature vector, and constructs a product vector library based on the product information feature vector; calculates the similarity between the business opportunity information vector and the product information feature vector in the vector database, and filters text or paragraphs with a similarity greater than a preset threshold as products that match the business opportunity information for recommendation. The present invention vectorizes business opportunity information to construct a business opportunity vector, vectorizes product information to construct a product information feature vector, constructs a product vector library based on the product information feature vector, calculates similarity to match products that best meet user needs, and thus improves the efficiency and accuracy of product recommendations. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] The accompanying drawings are only used to illustrate specific embodiments and are not considered to limit the present invention. In the entire drawings, the same reference symbols represent the same components. Obviously, the drawings described below are only some embodiments recorded in the embodiments of the present invention. For those of ordinary skill in the art, other drawings can also be obtained based on these drawings.

[0052] Figure 1 A method diagram for matching business opportunity information with products provided by an embodiment of the present invention;

[0053] Figure 2 A method diagram for constructing a feature word library provided by an embodiment of the present invention;

[0054] Figure 3 A method diagram for constructing a business opportunity vector provided by an embodiment of the present invention;

[0055] Figure 4 A method diagram for obtaining product information and preprocessing the product information provided by an embodiment of the present invention;

[0056] Figure 5 A method diagram for constructing a product vector library provided by an embodiment of the present invention;

[0057] Figure 6 A diagram of a device for implementing business opportunity information matching products provided by an embodiment of the present invention;

[0058] Figure 7 A schematic block diagram of an electronic device provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0059] In order to enable those skilled in the art to better understand the technical solutions in the embodiments of the present invention, the technical solutions of the present invention will be clearly and completely described below in conjunction with the accompanying drawings. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all of the embodiments. It should be understood that these descriptions are only exemplary and are not intended to limit the scope of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in the field without making creative work should fall within the scope of protection of the present invention.

[0060] Furthermore, in the following description, descriptions of well-known structures and technologies are omitted to avoid unnecessarily obscuring the concepts disclosed in the present invention.

[0061] In the description of the present invention, the terms "first", "second", and "third" are used for descriptive purposes only and should not be understood as indicating or implying relative importance. The terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be internal communication between two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to specific circumstances.

[0062] Exemplary embodiments will be described in detail herein, examples of which are shown in the accompanying drawings. When the following description refers to the drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present invention. Instead, they are merely examples of methods and systems consistent with some aspects of the present invention as detailed in the appended claims.

[0063] The present invention proposes a method, device, electronic device and storage medium for matching business opportunity information with products, vectorizes business opportunity information to construct a business opportunity vector, vectorizes product information to construct a product information feature vector, builds a product vector library based on the product information feature vector, calculates similarity to match products that best meet user needs, thereby improving the efficiency and accuracy of product recommendations.

[0064] Method Embodiment

[0065] Figure 1 A diagram of a method for matching business opportunity information with products provided in an embodiment of the present invention.

[0066] In a first aspect of the present invention, a method for matching business opportunity information with products is proposed. Figure 1 , including five steps from S1 to S5:

[0067] S1: Acquire business opportunity information, preprocess the business opportunity information, and build a feature vocabulary based on the preprocessed business opportunity information.

[0068] Specifically, in an embodiment of the present invention, a method for acquiring business opportunity information, preprocessing the business opportunity information, and constructing a feature word library based on the preprocessed business opportunity information is described in detail below.

[0069] S2: Vectorize each keyword in the feature vocabulary, set weights for each vectorized keyword and construct a business opportunity vector.

[0070] Specifically, in an embodiment of the present invention, each keyword in the feature word library is vectorized, and a method of setting a weight for each vectorized keyword to construct a business opportunity vector is described in detail below.

[0071] S3: Obtain product information and pre-process the product information.

[0072] Specifically, in the embodiment of the present invention, the method of obtaining product information and preprocessing the product information is described in detail below.

[0073] S4: Convert the preprocessed product information into a product information feature vector, and build a product vector library based on the product information feature vector.

[0074] Specifically, in an embodiment of the present invention, the preprocessed product information is converted into a product information feature vector, and a method for constructing a product vector library based on the product information feature vector is described in detail below.

[0075] S5: Calculate the similarity between the business opportunity information vector and the product information feature vector in the vector database, and filter out texts or paragraphs with similarity greater than a preset threshold as products matching the business opportunity information for recommendation.

[0076] Specifically, in an embodiment of the present invention, a method for calculating the similarity between a business opportunity information vector and a product information feature vector in a vector database, and screening texts or paragraphs with a similarity greater than a preset threshold to recommend products matching the business opportunity information is described in detail below.

[0077] Figure 2 A diagram of a method for constructing a feature word library provided in an embodiment of the present invention.

[0078] Furthermore, in the above method for matching business opportunity information with products, business opportunity information is obtained, the business opportunity information is preprocessed, a feature word library is constructed based on the preprocessed business opportunity information, and Figure 2 , including five steps from S21 to S25:

[0079] S21: Connect to the business opportunity system through an interface to collect key information to obtain business opportunity information;

[0080] S22: Clear invalid or duplicate data records in the business opportunity information;

[0081] S23: Performing a text segmentation operation, identifying and extracting key terms in the business opportunity information, performing part-of-speech tagging on the key terms, and screening the key terms after the part-of-speech tagging to obtain keywords;

[0082] S24: construct feature word library based on keywords;

[0083] S25: Classify the keywords in the feature word library according to the industry knowledge base to distinguish industry categories and demand types;

[0084] Among them, the key information includes at least: business opportunity title, customer needs, industry category and estimated amount.

[0085] Specifically, in an embodiment of the present invention, a business opportunity system is connected through an interface to collect key information including business opportunity title, customer demand, industry category, estimated amount, etc., and invalid or duplicate data records are cleared, and text segmentation operations are implemented to identify and extract key terms in the business opportunity description, such as industry identifiers and demand parameters; the identified key terms are tagged with parts of speech, and the tagged key terms are filtered to retain only nouns, verbs, adjectives, etc. A business opportunity feature vocabulary library is constructed based on keywords, and the keywords are classified according to the industry knowledge base to distinguish industry categories and demand types.

[0086] Figure 3 A diagram of a method for constructing a business opportunity vector provided by an embodiment of the present invention.

[0087] Furthermore, in the above method for matching business opportunity information to products, each keyword in the feature word library is vectorized, and a weight is set for each vectorized keyword to construct a business opportunity vector. Figure 3 , including two steps from S31 to S32:

[0088] S31: Use the pre-trained text embedding model to convert each keyword in the feature vocabulary into a vector representation;

[0089] S32: setting a weight for each vectorized keyword, and combining all feature vectors by weighted summation to obtain a business opportunity vector;

[0090] The weight is set based on the frequency of occurrence of keywords in the industry and their criticality in the business opportunity description.

[0091] Specifically, in an embodiment of the present invention, the pre-trained text embedding model can be BERT, GPT or Word2Vec. The pre-trained text embedding model is used to convert each keyword in the feature vocabulary into a vector representation. This conversion allows similarity calculation of business opportunity information in the vector space, and sets a corresponding weight for each keyword. The weight calculation can be based on the frequency of occurrence of the keyword in the industry and the criticality in the business opportunity description. Through the weighted summation method, all feature vectors are merged to obtain a comprehensive vector representing the entire business opportunity.

[0092] Figure 4 A diagram of a method for obtaining product information and preprocessing the product information provided by an embodiment of the present invention.

[0093] Furthermore, in the above method for matching business opportunity information with products, product information is obtained, the product information is preprocessed, and combined with Figure 4 ,include:

[0094] S41: extract product details from the software capability mall;

[0095] S42: Remove noise, duplicates and irrelevant information from product details;

[0096] S43: Determine the size of the document length of the product details after removing noise, duplicates and irrelevant information and the size of the preset threshold;

[0097] S44: If the determination result is that the document length of the product details is greater than a preset threshold, the document is segmented into a plurality of text blocks;

[0098] Among them, product details include at least: product name, description, industry and sales data.

[0099] Specifically, in an embodiment of the present invention, product details are extracted from the software capability mall, involving information such as product name, description, industry, sales data, etc.; the product details are noise-removed, duplicated, and irrelevant information is processed to ensure the quality and accuracy of the data. In order to reduce the burden of the pre-trained text embedding model and improve the matching accuracy, long documents are segmented into multiple text blocks.

[0100] Figure 5 A diagram of a method for constructing a product vector library provided in an embodiment of the present invention.

[0101] Furthermore, in the above method for matching business opportunity information with products, the preprocessed product information is converted into a product information feature vector, a product vector library is constructed based on the product information feature vector, and a product vector library is constructed based on the product information feature vector. Figure 5 , including five steps from S51 to S55:

[0102] S51: Perform word segmentation and keyword extraction on the pre-processed product information to obtain product information features.

[0103] S52: Use the pre-trained text embedding model to convert product information features into vector representation to obtain a product information feature vector.

[0104] Specifically, in an embodiment of the present invention, the pre-trained text embedding model can be BERT, GPT or Word2Vec, and the product information features are converted into a vector representation to obtain a product information feature vector. This conversion enables similarity calculation of product information in the vector space.

[0105] Here, the pre-trained text embedding model used for product information feature vectorization and business opportunity information vectorization is the same.

[0106] S53: Construct a dictionary based on the product information feature vector, and map each word in the dictionary to an integer to obtain a dictionary index.

[0107] Specifically, in an embodiment of the present invention, a dictionary is constructed based on the product information feature vector, and its capacity is represented by |V|. Each word in the dictionary is uniquely mapped to an integer from 0 to |V|-1, and these integers are designated as dictionary indexes. Based on this, a dictionary index set V is formed, which is formally represented as V={0,1,...,|V|-1}. In this set, each word is assigned a specific index value, which is the unique identifier of the word.

[0108] S54: For each word with dictionary index i in the dictionary, two sets of word vector representations are obtained when it is a central word and a background word respectively.

[0109] Specifically, in the embodiment of the present invention, for each word with a dictionary index i in the dictionary, the method for obtaining two sets of word vector representations when it is a central word and a background word is as follows:

[0110] For a text sequence of length T, set the time step t, the vocabulary at time t is recorded as w(t). When the background window size is n, the skip-word model needs to maximize the probability of generating background words given any central word, which is expressed as:

[0111]

[0112] The maximum likelihood estimation obtained from the above formula is equivalent to minimizing the following loss function

[0113]

[0114] Define v as the word vector representation of the center word, and u as the word vector representation of the background word. In order to integrate the learning objective of the model into the loss function, the model parameters, that is, these word vectors, are used to explicitly express the probability of the center word generating its surrounding background words in the loss function. This probability assumes that when the center word is given, the appearance of each background word is independent of each other, regardless of the correlation between the background words. Given the center word W c The index in the dictionary is c, the background word W o The index in the dictionary is o, and the probability of the center word generating the background word in the loss function can be defined using the softmax function:

[0115]

[0116] When the sequence length T is large, a smaller subsequence is randomly sampled to calculate the loss function and SGD is used to optimize the loss function. By taking the derivative, the logarithm of the probability of generating the above formula can be calculated with respect to the center word vector V C The gradient of is:

[0117]

[0118] The first part of the derivation:

[0119]

[0120] The second part of the derivation:

[0121]

[0122] In summary:

[0123]

[0124] After obtaining the gradient information calculated above, the stochastic gradient descent SGD algorithm is used as the optimization strategy to iteratively update the model parameters v c , that is, the word vector of the central word. Similarly, for other model parameters u o That is, the word vector of the background word also follows the same iterative update principle. This process continues until for each word with index i in the dictionary, two sets of word vector representations can be obtained when it is the center word and the background word, that is, v i and u i ;

[0125] S55: Build a product vector library based on two sets of word vectors.

[0126] Specifically, in the embodiment of the present invention, based on v i and u i Build a product vector library.

[0127] Furthermore, in the above method for matching business opportunity information with products, the similarity between the business opportunity information vector and the product information feature vector in the vector database is calculated by cosine similarity.

[0128] Specifically, in the embodiment of the present invention, the vectors are regarded as arrows starting from the origin, and the cosine similarity is the cosine value of the angle between the two arrows. This value is between -1 and 1. The larger the value, the more similar the two vectors are, and the smaller the value, the less similar the two vectors are.

[0129] In two-dimensional space, the calculation formula of cosine similarity is expressed as cos(θ)=(A·B) / (||A||||B||), where A and B are two vectors, ||A|| and ||B|| represent the modulus (length) of A and B respectively, and A·B represents the dot product of A and B.

[0130] In multidimensional space, let A(x1,y1,z1,…) and B(x2,y2,z2,…), then the cosine similarity is: (Σ(A i B i)) / (||A||||B||), where Σ represents the sum of all components, and A(x1,y1,z1,…) can be a business opportunity vector. Similarly, B(x2,y2,z2,…) is the feature vector of product information in the vector database. The specific cos(θ) can be calculated through cosine similarity, thereby obtaining the similarity.

[0131] Furthermore, in the above method for matching business opportunity information with products, the text or paragraphs with a similarity greater than a preset threshold are screened and recommended as products matching the business opportunity information, including:

[0132] The text or paragraphs with similarity greater than a preset threshold are taken as preliminary matching results;

[0133] The preliminary matching results are arranged from high to low according to the similarity score, and the text or paragraph with the highest similarity score is recommended as the product matching the business opportunity information.

[0134] Specifically, in the embodiment of the present invention, the text or paragraph with the highest similarity score is recommended as a product matching the business opportunity information, thereby improving the efficiency and accuracy of product recommendation.

[0135] Device Embodiment

[0136] Figure 6 A diagram of a device for implementing business opportunity information matching products provided by an embodiment of the present invention.

[0137] In a second aspect of the present invention, a device for matching business opportunity information with products is also provided. Figure 6 ,include:

[0138] The acquisition module 61 is used to acquire business opportunity information, pre-process the business opportunity information, and construct a feature word library based on the pre-processed business opportunity information.

[0139] Specifically, in the embodiment of the present invention, the acquisition module 61 acquires business opportunity information, preprocesses the business opportunity information, and constructs a feature word library based on the preprocessed business opportunity information.

[0140] Construction module 62: used to vectorize each keyword in the feature word library, set a weight for each vectorized keyword and construct a business opportunity vector.

[0141] Specifically, in the embodiment of the present invention, the construction module 62 vectorizes each keyword in the feature word library, sets a weight for each vectorized keyword, and constructs a business opportunity vector.

[0142] Preprocessing module 63: used to obtain product information and preprocess the product information.

[0143] Specifically, in the embodiment of the present invention, the preprocessing module 63 obtains product information and preprocesses the product information.

[0144] Conversion module 64: used to convert the pre-processed product information into a product information feature vector, and build a product vector library based on the product information feature vector.

[0145] Specifically, in the embodiment of the present invention, the preprocessing module 63 converts the preprocessed product information into a product information feature vector, and constructs a product vector library based on the product information feature vector.

[0146] Matching module 65: used to calculate the similarity between the business opportunity information vector and the product information feature vector in the vector database, and filter the text or paragraphs with similarity greater than a preset threshold as products matching the business opportunity information for recommendation.

[0147] Specifically, in the embodiment of the present invention, the matching module 65 calculates the similarity between the business opportunity information vector and the product information feature vector in the vector database, and selects texts or paragraphs with similarity greater than a preset threshold value as products matching the business opportunity information for recommendation.

[0148] According to a third aspect of the present invention, an electronic device is provided, comprising: a processor and a memory;

[0149] The processor is used to execute any one of the above methods for matching business opportunity information with products by calling the program or instruction stored in the memory.

[0150] In a fourth aspect of the present invention, a computer-readable storage medium is provided, wherein the computer-readable storage medium stores a program or instruction, and the program or instruction enables a computer to execute any one of the above methods for matching business opportunity information with products.

[0151] Figure 7 It is a schematic block diagram of an electronic device provided by an embodiment of the present invention.

[0152] like Figure 7 As shown, the electronic device includes: at least one processor 701, at least one memory 702 and at least one communication interface 703. The various components in the electronic device are coupled together through a bus system 704. The communication interface 703 is used to transmit information between external devices. It can be understood that the bus system 704 is used to achieve connection and communication between these components. In addition to the data bus, the bus system 704 also includes a power bus, a control bus and a status signal bus. However, for the sake of clarity, Figure 7 Various buses are labeled as bus system 704 .

[0153] It can be understood that the memory 702 in this embodiment can be a volatile memory or a non-volatile memory, or can include both volatile and non-volatile memories.

[0154] In some implementations, the memory 702 stores the following elements, executable units or data structures, or a subset thereof, or an extended set thereof: an operating system and application programs.

[0155] The operating system includes various system programs, such as a framework layer, a core library layer, a driver layer, etc., which are used to implement various basic services and process hardware-based tasks. The application includes various application programs, such as a media player (Media Player), a browser (Browser), etc., which are used to implement various application services. A program that implements any method of a method for matching business opportunity information with products provided in an embodiment of the present invention can be included in the application.

[0156] In an embodiment of the present invention, the processor 701 calls a program or instruction stored in the memory 702, specifically, a program or instruction stored in an application, and the processor 701 is used to execute the steps of each embodiment of a method for matching business opportunity information with products provided in an embodiment of the present invention.

[0157] Acquire business opportunity information, pre-process the business opportunity information, and build a feature word library based on the pre-processed business opportunity information;

[0158] Vectorize each keyword in the feature word library, set weights for each vectorized keyword and construct a business opportunity vector;

[0159] Obtain product information and pre-process the product information;

[0160] Convert the preprocessed product information into a product information feature vector, and build a product vector library based on the product information feature vector;

[0161] The similarity between the business opportunity information vector and the product information feature vector in the vector database is calculated, and the text or paragraphs with similarity greater than a preset threshold are filtered out as products matching the business opportunity information for recommendation.

[0162] Any method in a method for matching business opportunity information with products provided in an embodiment of the present invention can be applied to the processor 701, or implemented by the processor 701. The processor 701 can be an integrated circuit chip with signal processing capabilities. In the implementation process, each step of the above method can be completed by an integrated logic circuit of hardware in the processor 701 or an instruction in the form of software. The above processor 701 can be a general-purpose processor, a digital signal processor (DigitalSignal Processor, DSP), an application-specific integrated circuit (Application Specific Integrated Circuit, ASIC), a field programmable gate array (Field Programmable Gate Array, FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc.

[0163] The steps of any method in the method for matching business opportunity information with products provided in the embodiment of the present invention can be directly embodied as being executed by a hardware decoding processor, or being executed by a combination of hardware and software units in the decoding processor. The software unit can be located in a mature storage medium in the art, such as a random access memory, a flash memory, a read-only memory, a programmable read-only memory, or an electrically erasable programmable memory, a register, etc. The storage medium is located in the memory 702, and the processor 701 reads the information in the memory 702 and completes the steps of the method in combination with its hardware.

[0164] Those skilled in the art will appreciate that although some embodiments described herein include certain features included in other embodiments but not other features, the combination of features from different embodiments is meant to be within the scope of the present invention and to form different embodiments.

[0165] Those skilled in the art will appreciate that the description of each embodiment has its own emphasis, and for parts not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0166] Although the embodiments of the present invention are described in conjunction with the accompanying drawings, those skilled in the art may make various modifications and variations without departing from the spirit and scope of the present invention, and such modifications and variations all fall within the scope defined by the appended claims. The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art who is familiar with the technical field can easily think of various equivalent modifications or substitutions within the technical scope disclosed by the present invention, and these modifications or substitutions should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.

[0167] The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A method for matching business opportunity information with products, characterized in that: include: Acquire business opportunity information, preprocess the business opportunity information, and build a feature word library based on the preprocessed business opportunity information; Vectorize each keyword in the feature word library, set weights for each vectorized keyword and construct a business opportunity vector; Acquire product information, and pre-process the product information; Convert the preprocessed product information into a product information feature vector, and build a product vector library based on the product information feature vector; The similarity between the business opportunity information vector and the product information feature vector in the vector database is calculated, and the text or paragraphs with similarity greater than a preset threshold are filtered out as products matching the business opportunity information for recommendation.

2. A method for matching business opportunity information with products according to claim 1, characterized in that: Acquiring business opportunity information, preprocessing the business opportunity information, and building a feature word library based on the preprocessed business opportunity information, including: Connect to the business opportunity system through the interface to collect key information and obtain business opportunity information; Clear invalid or duplicate data records in business opportunity information; Implement text segmentation operations, identify and extract key terms in business opportunity information, perform part-of-speech tagging on key terms, and filter the tagged key terms to obtain keywords; Build a feature vocabulary based on keywords; Classify the keywords in the feature word library according to the industry knowledge base to distinguish industry categories and demand types; Among them, the key information includes at least: business opportunity title, customer needs, industry category and estimated amount.

3. A method for matching business opportunity information with products according to claim 1, characterized in that: Each keyword in the feature word library is vectorized, and a weight is set for each vectorized keyword to construct a business opportunity vector, including: Use a pre-trained text embedding model to convert each keyword in the feature vocabulary into a vector representation; Set a weight for each vectorized keyword, and combine all feature vectors to obtain the business opportunity vector by weighted summation; The weight is set based on the frequency of occurrence of keywords in the industry and their criticality in the business opportunity description.

4. A method for matching business opportunity information with products according to claim 1, characterized in that: Acquire product information and pre-process the product information, including: Extract product details from the software capability mall; Remove noise, duplicates and irrelevant information from product details; Determine the size of the document length of the product details after removing noise, duplicates and irrelevant information and the size of the preset threshold; If the judgment result is that the document length of the product details is greater than the preset threshold, the document is divided into multiple text blocks; Among them, product details include at least: product name, description, industry and sales data.

5. A method for matching business opportunity information with products according to claim 1, characterized in that: The preprocessed product information is converted into a product information feature vector, and a product vector library is constructed based on the product information feature vector, including: Perform word segmentation and keyword extraction on the pre-processed product information to obtain product information features; Use the pre-trained text embedding model to convert product information features into vector representation to obtain product information feature vector; Build a dictionary based on the product information feature vector, and map each word in the dictionary to an integer to get the dictionary index; For each word with dictionary index i in the dictionary, obtain two sets of word vector representations when it is the center word and background word respectively; Build a product vector library based on two sets of word vectors; Here, i represents an integer.

6. A method for matching business opportunity information with products according to claim 1, characterized in that: The similarity between the business opportunity information vector and the product information feature vector in the vector database is calculated by cosine similarity.

7. A method for matching business opportunity information with products according to claim 1, characterized in that: The step of screening texts or paragraphs with a similarity greater than a preset threshold as products matching the business opportunity information for recommendation includes: The text or paragraphs with similarity greater than a preset threshold are taken as preliminary matching results; The preliminary matching results are arranged from high to low according to the similarity score, and the text or paragraph with the highest similarity score is recommended as the product matching the business opportunity information.

8. A device for matching business opportunity information with products, characterized in that: include: Acquisition module: used to acquire business opportunity information, preprocess the business opportunity information, and build a feature word library based on the preprocessed business opportunity information; Construction module: used to vectorize each keyword in the feature vocabulary, set weights for each vectorized keyword and construct business opportunity vectors; Preprocessing module: used to obtain product information and preprocess the product information; Conversion module: used to convert the preprocessed product information into product information feature vectors, and build a product vector library based on the product information feature vectors; Matching module: used to calculate the similarity between the business opportunity information vector and the product information feature vector in the vector database, and filter the text or paragraphs with similarity greater than a preset threshold as products matching the business opportunity information for recommendation.

9. An electronic device, characterized in that: include: Processor and memory; The processor is used to execute the method for matching business opportunity information with products as described in any one of claims 1 to 7 by calling the program or instruction stored in the memory.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a program or instruction, and the program or instruction enables a computer to execute a method for matching business opportunity information with products as claimed in any one of claims 1 to 7.

Images