A quick search method and system for querying similar enterprises

Abstract

A kind of fast retrieval method and system for querying similar enterprises, it is related to the technical field of enterprise query.The method application server, the method includes: obtaining the main business of target enterprise input by user;First characterization vector is obtained by matching main business with preset vector representation database, and first characterization vector includes first enterprise cluster number;First enterprise cluster number is matched with preset enterprise cluster library, and the cluster cluster corresponding to first enterprise cluster number is obtained, and the cluster cluster corresponding to first enterprise cluster number includes the main business document of multiple enterprises with the same main business type as target enterprise;Based on the cluster cluster corresponding to the first enterprise cluster number, obtain multiple enterprises with the same main business type as the target enterprise.The problem that the recall document relevance is not high enough is solved by the fact that the currently commonly used retrieval method only considers the similarity on the basis of word segmentation.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of enterprise query, in particular to a quick retrieval method and system for querying similar enterprises. BACKGROUND

[0002] With the increase of the number of enterprises, and the difference of the main business of each enterprise, in the process of mining similar enterprises, the accuracy of the representation of the main business of the enterprise and the huge calculation of the similarity between each main business representation need to be considered. Therefore, how to construct the enterprise peer portrait through a quick and effective similar enterprise retrieval method plays a key role in mining similar enterprises.

[0003] At present, in the information retrieval process, the commonly used retrieval method is ES retrieval (Elasticsearch, Elasticsearch). It obtains a plurality of feature words by segmenting the content to be retrieved, and then calculates the similarity between the plurality of feature words after segmentation and the enterprise database to recall the documents related to the content to be retrieved.

[0004] However, the above retrieval method only considers the similarity of the segmented words in the literal sense, which leads to insufficient relevance of the recalled documents. SUMMARY

[0005] In view of the problem that the commonly used retrieval method only considers the similarity of the segmented words in the literal sense, which leads to insufficient relevance of the recalled documents, the present application provides a quick retrieval method and system for querying similar enterprises.

[0006] In a first aspect, the present application provides a quick retrieval method for querying similar enterprises, applied to a server, the method comprising: obtaining the main business of a target enterprise input by a user; matching the main business with a preset vector representation database to obtain a first representation vector, the first representation vector comprising a first enterprise cluster number, the first enterprise cluster number representing the vector number of the main business type of the target enterprise, the preset vector representation database comprising the corresponding relationship between the main business and its corresponding representation vector; matching the first enterprise cluster number with a preset enterprise cluster library to obtain a cluster cluster corresponding to the first enterprise cluster number, the preset enterprise cluster library comprising the corresponding relationship between the cluster number and the cluster cluster, the cluster cluster corresponding to the first enterprise cluster number comprising the main business documents of a plurality of enterprises with the same main business type as the target enterprise; and obtaining a plurality of enterprises with the same main business type as the target enterprise based on the cluster cluster corresponding to the first enterprise cluster number.

[0007] By adopting the technical scheme, the main business of the target enterprise is matched with the preset vector representation database, so that the first representation vector with high representation capability is obtained, thereby solving the problem of low vector representation capability of the main business, and the first enterprise clustering number in the first representation vector is matched with the clustering cluster corresponding to the first enterprise clustering number in the preset enterprise clustering library, so that the main business of the target enterprise does not need to be matched with the enterprise database again, and only by the clustering number corresponding to the clustering cluster, other enterprises with high similarity to the main business of the target enterprise can be obtained, thereby improving the retrieval efficiency.

[0008] In a second aspect, the application provides a fast retrieval system for querying similar enterprises, the system being a server, the server comprising an acquisition module and a processing module, wherein:

[0009] The acquisition module is configured to acquire the main business of a target enterprise input by a user.

[0010] The processing module is configured to match the main business with a preset vector representation database to obtain a first representation vector, the first representation vector comprising a first enterprise clustering number, the first enterprise clustering number representing a vector number of a main business type of the target enterprise, the preset vector representation database comprising a corresponding relationship between a main business and a representation vector corresponding to the main business; match the first enterprise clustering number with a preset enterprise clustering library to obtain a clustering cluster corresponding to the first enterprise clustering number, the preset enterprise clustering library comprising a corresponding relationship between a clustering number and a clustering cluster, the clustering cluster corresponding to the first enterprise clustering number comprising main business documents of a plurality of enterprises with the same main business type as the target enterprise; and obtain the plurality of enterprises with the same main business type as the target enterprise based on the clustering cluster corresponding to the first enterprise clustering number.

[0011] Optionally, before matching the main business with the preset vector representation database to obtain the first representation vector, the following steps are included:

[0012] The acquisition module acquires a plurality of brief introduction texts of enterprises; and extracts main businesses in the plurality of brief introduction texts by adopting a preset rule screening model, the preset rule screening model being defined with a plurality of information screening rules, wherein one brief introduction text corresponds to one main business, and one main business comprises a plurality of main business words.

[0013] The processing module inputs the first main business into a plurality of preset pre-training models to obtain a plurality of first text vectors, the first main business being any one of the plurality of main businesses; performs first pooling processing on the plurality of first text vectors respectively to generate a plurality of second text vectors; performs second pooling processing on the plurality of second text vectors to obtain a third text vector; and stores the third text vector as a representation vector corresponding to the first main business in the preset vector representation database.

[0014] By adopting the technical scheme, the first pooling processing is performed on the main business of the enterprise, and vectors of a plurality of main business words contained in the main business are fused into a comprehensive vector, thereby reducing the data operation amount; then the comprehensive vector is input into a plurality of pre-training models, and the output result is subjected to second pooling, thereby reducing the decline degree of the vector representation accuracy of the main business caused by the first pooling processing, and the better representation capability of the plurality of pre-training models is fused, and the vector representation of the main business is greatly improved.

[0015] Optionally, before the main business is matched with the preset vector representation database to obtain the first representation vector, the method further includes: a obtaining module calling a plurality of word vector models from a preset word vector model library; constructing a plurality of main businesses of the enterprise into a training sample library; a processing module inputting data in the training sample library into the plurality of word vector models to obtain a plurality of training text vectors; calculating respective evaluation values of the plurality of word vector models based on the plurality of training text vectors by using a vector space distribution evaluation function; sorting the respective evaluation values of the plurality of word vector models from large to small to generate a sorting table; and selecting a plurality of word vector models with a ranking greater than or equal to a preset ranking in the sorting table as pre-training models.

[0016] By adopting the technical scheme, since the representation capabilities of different pre-training models are different, and a word vector model with a better representation capability needs a large amount of time for retraining. At this time, the plurality of trained word vector models are directly called as pre-training models, then the training sample library storing the main business is input into the plurality of word vector models to evaluate the representation capabilities of the plurality of word vector models, and several word vector models with better representation capabilities are selected as pre-training models according to the evaluation values of the plurality of word vector models. Thus, a large amount of time is no longer needed for training the model, and the plurality of pre-training models with better representation capabilities are obtained.

[0017] Optionally, based on the plurality of training text vectors, the respective evaluation values of the plurality of word vector models are calculated by using a vector space distribution evaluation function, and the calculation is specifically as follows:

[0018]

[0019] wherein, T is the evaluation value, W xi is the training text vector corresponding to the i th main business word of the x th main business, W yj is the training text vector corresponding to the i th main business word of the y th main business, n is the number of main business categories, m is the number of main business words in the n th main business category, the numerator is used to represent the average similarity between the plurality of main business words of the same main business, and the denominator is used to represent the average similarity value between the plurality of main business words corresponding to different main businesses.

[0020] By adopting the above technical solution, based on the calculation formula of the evaluation value of the word vector model, whether the word vector model can associate multiple main words of the same main business and whether the word vector model can distinguish multiple main words corresponding to different main businesses, based on the evaluation value, it is determined whether the word vector model meets the representation requirements of the pre-training model.

[0021] Optionally, the processing module traverses the word vector dimensions of the plurality of word vector models; if the first word vector dimension is less than or equal to the second word vector dimension, the first word vector dimension is determined as the standard vector dimension, and the first word vector dimension and the second word vector dimension are any two word vector dimensions in the plurality of word vector dimensions; and the word vector dimensions corresponding to the plurality of word vector models are respectively compressed to the standard vector dimension.

[0022] By adopting the above technical solution, the word vector dimensions of the plurality of word vector models are compressed to a unified vector dimension, the word vectors output by the plurality of word vector models are aligned in the vector space, and the output results of the plurality of word vector models can be fused.

[0023] Optionally, before matching the first enterprise clustering number with the preset enterprise clustering library to obtain the clustering cluster corresponding to the first enterprise clustering number, the obtaining module determines a plurality of enterprise clustering numbers of the preset enterprise clustering library based on the main business classification tree; the processing module clusters the representation vectors of the plurality of main businesses in the preset vector representation database based on the plurality of enterprise clustering numbers to obtain a plurality of clustering clusters, wherein one enterprise clustering number corresponds to one clustering cluster; and the plurality of enterprise clustering numbers and the clustering clusters corresponding to the plurality of enterprise clustering numbers are constructed as a corresponding relationship and stored in the preset enterprise clustering library.

[0024] By adopting the above technical solution, since the number of main businesses is large, the clustering cluster classifies and summarizes the huge main business data, gathers the main business data with high similarity to the same clustering cluster, and divides the main business data with low similarity to different clustering clusters, and finally represents one clustering cluster with one enterprise clustering number, thereby reducing the data processing amount during data matching and improving the efficiency of data matching.

[0025] Optionally, the obtaining module obtains the representation vector of the second main business and the center vector of the plurality of enterprise clustering numbers, and the second main business is any one of the plurality of main businesses in the preset vector representation database; when the cosine distance value between the representation vector of the second main business and the center vector of the second enterprise clustering number is greater than or equal to a preset threshold, the processing module constructs the representation vector of the second main business and the second enterprise clustering number as a corresponding relationship, and the second enterprise clustering number is any one of the plurality of enterprise clustering numbers.

[0026] By adopting the technical solution, the cosine distance value between the second main business representation vector and the center vector of the plurality of enterprise cluster numbers is determined, so as to determine which cluster corresponding to the enterprise cluster number is most similar to the second main business, thereby improving the relevance between the same type of main businesses, and distinguishing different types of main businesses, so that the recall result corresponding to the main business of the target enterprise has high relevance.

[0027] In a third aspect, the present application provides an electronic device, comprising a processor, a memory, a user interface and a network interface, the memory is used to store instructions, the user interface and the network interface are used to communicate with other devices, and the processor is used to execute the instructions stored in the memory, so that the electronic device executes the method according to any one of the first aspect.

[0028] In a fourth aspect, the present application provides a computer readable storage medium, which stores instructions, when the instructions are executed, the method according to any one of the first aspect is executed.

[0029] In summary, the one or more technical solutions provided in the embodiments of the present application have at least the following technical effects or advantages:

[0030] 1. The main business of the target enterprise is matched with the preset vector representation database, so as to obtain the first representation vector with high representation ability; so as to solve the problem of low vector representation ability of the main business, and then the first enterprise cluster number in the first representation vector is matched with the cluster corresponding to the first enterprise cluster number in the preset enterprise cluster library, at this time, it is not necessary to traverse and match the main business of the target enterprise with the enterprise database, and only through the cluster number corresponding to the cluster, the main business documents of other enterprises with high similarity to the main business of the target enterprise can be obtained, thereby improving the efficiency of retrieval.

[0031] 2. Since the representation abilities of different pre-training models are different, and it takes a lot of time to retrain a word vector model with good representation ability. At this time, a plurality of trained word vector models are directly called as pre-training models, and then the training sample library storing the main business is input into the plurality of word vector models, the representation abilities of the plurality of word vector models are evaluated, and several word vector models with good representation ability are selected as pre-training models according to the evaluation values of the plurality of word vector models. Therefore, it is no longer necessary to spend a lot of time to train the model, and a plurality of pre-training models with good representation ability are obtained. BRIEF DESCRIPTION OF DRAWINGS

[0032] Figure 1 is a flowchart of a fast retrieval method for querying similar enterprises provided by the embodiments of the present application.

[0033] Figure 2is a scene schematic diagram of a first pooling processing provided by an embodiment of the present application.

[0034] Figure 3 is a structure schematic diagram of a fast retrieval system for querying similar enterprises provided by an embodiment of the present application.

[0035] Figure 4 is a structure schematic diagram of an electronic device provided by an embodiment of the present application.

[0036] Legend: 1, acquisition module; 2, processing module; 400, electronic device; 401, processor; 402, communication bus; 403, user interface; 404, network interface; 405, memory. DETAILED DESCRIPTION

[0037] In order for those skilled in the art to better understand the technical solutions in the specification, the technical solutions in the specification will be clearly and completely described below in conjunction with the drawings in the specification. Obviously, the described embodiments are only some of the embodiments of the present application, not all the embodiments.

[0038] In the description of the embodiments of the present application, the words such as "for example" or "for instance" are used to represent an example, illustration or description. Any embodiment or design scheme described as "for example" or "for instance" in the embodiments of the present application should not be interpreted as more preferred or more advantageous than other embodiments or design schemes. Rather, the words such as "for example" or "for instance" are intended to present the relevant concept in a specific manner.

[0039] In the description of the embodiments of the present application, the term "a plurality of" means two or more. For example, a plurality of systems means two or more systems, and a plurality of screen terminals means two or more screen terminals. In addition, the terms "first" and "second" are used for description purposes only, and should not be interpreted as indicating or implying relative importance or implicitly indicating the indicated technical features. Therefore, the features defined with "first" and "second" can explicitly or implicitly include one or more features. The terms "include", "contain", "have" and their variants mean "include but are not limited to", unless otherwise specifically emphasized.

[0040] With the development of social economy, the development of enterprises gradually develops towards diversification. Even if the industry types of two enterprises are the same, there are still differences in their main businesses. Therefore, when mining similar enterprises of an enterprise, the representation of the main business of the enterprise is very important, and the accuracy of the representation will directly affect the result of mining similar enterprises.

[0041] The currently commonly used information retrieval method is ES retrieval (Elasticsearch). The ES retrieval obtains multiple feature words by segmenting the retrieval information input by the user, and then calculates the similarity between the multiple feature words and the enterprise database to recall the documents related to the retrieval information. However, the ES retrieval does not perform much representation processing on the representation of the retrieval information before retrieval. Only after the multiple feature words are converted into a text vector, the similarity calculation with the enterprise database is performed, which is prone to cause the situation that the recall result is not strongly associated with the retrieval information. The representation processing can be understood as normalizing and adjusting the information to express its accurate meaning. In addition, in the retrieval process, the ES retrieval mainly searches for the associated documents in the enterprise database by traversing and comparing the multiple feature words with all the data in the enterprise database. This process has a large amount of calculation, thereby resulting in low retrieval efficiency.

[0042] To solve the above problems, the present application provides a fast retrieval method for querying similar enterprises. The method is applied to a server. The server can be an independent physical server, a server cluster or a distributed system composed of multiple physical servers, a cloud server providing cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, content distribution networks (CDN), and big data and artificial intelligence platforms, etc. The present disclosure does not limit the basic cloud computing services. As shown in the figure, the method comprises steps S101 to S104. Figure 1

[0043] S101, obtaining the main business of the target enterprise input by the user.

[0044] In the above steps, the user can input the main business of the target enterprise through the enterprise information query application installed on the user terminal. At this time, the user terminal sends a query request to the server and sends the user input content to the server. The user device can be a mobile phone, a tablet computer, a desktop computer, a personal computer, etc. without limitation. The user input information includes but is not limited to text input, voice input, historical browsing record input, etc.

[0045] S102, matching the main business with a preset vector representation database to obtain a first representation vector, the first representation vector comprising a first enterprise cluster number, the first enterprise cluster number representing the vector number of the main business type of the target enterprise, and the preset vector representation database comprising the corresponding relationship between the main business and its corresponding representation vector.

[0046] ​In the above steps, the user-inputted main business of the target company is not an accurate description. Therefore, to ensure the accuracy and efficiency of the main business representation, a vector representation database is pre-built, storing the correspondence between main businesses and their corresponding representation vectors. The vector representation database is constructed as follows: First, the server obtains a large amount of introductory text about companies; then, a rule-based filtering model is pre-defined, containing multiple information filtering rules. These rules can be understood as the content between, before, and after pre-labeled filtering terms. For example, "dedicated to" XX "field", XX "solutions", and "under its umbrella" XX can be considered a set of information filtering rules. Through these multiple information filtering rules defined in the model, the main businesses of each company can be extracted from a large amount of introductory text.

[0047] For a company, its main business is often more than one, thus a selected main business may include multiple key terms. Taking the first main business as an example, the first main business can be the main business of any one of multiple companies, and it includes multiple key terms. Then, these multiple key terms are treated as a dataset and input into several pre-trained models to obtain the representation vectors of the first main business in different pre-trained models. These are the multiple first text vectors corresponding to the first main business, where each first text vector is composed of the representation vectors of each key term. The pre-trained model can be understood as a word vector model, used to convert text into numerical vectors. Since different pre-trained models represent text into vectors with different dimensions, to reduce the computational load on the server, the first text vectors are... The representation vectors of multiple main keywords in the data are combined into a low-dimensional composite vector, i.e., the second text vector. At this time, the second text vector only includes a vector of one dimension. The specific process is as follows: the first text vector is subjected to a first pooling process to generate the second text vector. The first pooling process can be understood as reducing the dimensionality of the first text vector. The first pooling process is specifically: using the main keyword evaluation model to obtain the weight values ​​corresponding to each of the multiple main keywords in the first main business, mapping the weight values ​​to the multiple dimensions of data in the first text vector, and then performing a weighted average of the multiple dimensions of data to obtain the composite vector of the multiple dimensions of data, i.e., the second text vector. At this time, the second text vector represents the meaning of all the main keywords in the first main business.

[0048] For example, such as Figure 2As shown, the first main business includes four main words (w1, w2, w3, w4), and the pre-training model includes a No. 1 pre-training model, a No. 2 pre-training model, and a No. 3 pre-training model. The first main business is input into the three pre-training models respectively, and three pre-training models respectively correspond to the first text vectors [o1, o2, o3, o4], [p1, p2, p3, p4], and [q1, q2, q3, q4] are obtained. The three representation vectors are subjected to the first pooling processing, and the second text vectors [O], [P], and [Q] corresponding to the three representation vectors are obtained.

[0049] However, the first pooling processing will cause the representation ability of the second text vector to the first main business to decrease, and therefore, multiple pre-training models with better representation ability need to be selected to represent the first main business, so as to alleviate the representation precision decrease caused by the first pooling processing. The specific process is as follows: multiple word vector models are called from a preset word vector model library, and the preset word vector model library includes multiple trained word vector models. Then, the main businesses of multiple enterprises are constructed into a training sample library, and the data in the training sample library are input into the multiple word vector models to obtain multiple training text vectors, wherein one main business corresponds to multiple training text vectors. Then, the representation ability of the multiple word vector models is evaluated according to the multiple training text vectors, and the evaluation values of the multiple word vector models are obtained, and the calculation method adopts a vector space distribution evaluation function, and the specific calculation method is as follows:

[0050]

[0051] wherein, T is the evaluation value, W xi is the training text vector corresponding to the i-th main word of the x-th main business, W yj is the training text vector corresponding to the i-th main word of the y-th main business, n is the number of main business categories, m is the number of main words in the n-th main business category, the numerator is used to represent the average similarity between the multiple main words of the same main business, and the denominator is used to represent the average similarity value between the multiple main words corresponding to different main businesses.

[0052] Finally, the evaluation values of the multiple word vector models are sorted in descending order to generate a sorting table, and the multiple word vector models with a ranking greater than or equal to a preset ranking in the sorting table are selected as the pre-training models. At this time, the first main business is represented by the multiple pre-training models with better representation ability, so as to reduce the precision decrease problem caused by the dimension reduction of the second text vector.

[0053] Since the representation capabilities of the plurality of pre-training models are different, the representation capabilities of the plurality of pre-training models on the first main business are fused by performing second pooling processing on the plurality of second text vectors corresponding to the first main business, so as to further improve the representation capability of the representation vector of the first main business. However, different pre-training models have different dimensions of representing the main business, and the data dimensions of the plurality of pre-training models need to be unified. At this time, by traversing the word vector dimensions of the plurality of word vector models, in order to reduce the processing amount of data, the smallest word vector dimension in the plurality of word vector dimensions is selected as the standard vector dimension. Taking the first word vector dimension and the second word vector dimension as an example, any two word vector dimensions in the plurality of word vector dimensions, if the first word vector dimension is less than or equal to the second word vector dimension, the first word vector dimension is determined as the standard vector dimension, and the word vector dimensions of the plurality of word vector models are compressed to the standard vector dimension. Of course, the standard vector dimension is not limited to the smallest word vector dimension. The standard vector dimension is the vector dimension that can meet the minimum representation requirement of the representation capability of the main business. Therefore, the standard vector dimension can be any one of the word vector dimensions between the smallest word vector dimension and the largest word vector dimension. It needs to be explained that the more the vector dimensions of the standard vector dimension, the more accurate the representation capability, but the higher the corresponding computing power is required. Therefore, when selecting the plurality of pre-training models, the word vector dimensions of the plurality of pre-training models meet the minimum representation requirement, and the smallest word vector dimension is preferably selected as the standard vector dimension, that is, the accuracy of the representation of the word vector model can be ensured, and the data processing efficiency can be improved.

[0054] The second pooling processing is specifically: performing weighted average on the plurality of second text vectors corresponding to the first main business to obtain a third text vector, and storing the third text vector as a representation vector corresponding to the first main business in a preset vector representation database.

[0055] When the preset vector representation database is used, the main business of the target enterprise is matched with the preset vector representation database, and according to the corresponding relationship between the representation vector and the main business, a first representation vector corresponding to the main business of the target enterprise is obtained. The first representation vector includes a first enterprise cluster number, and the first enterprise cluster number is a specific vector data, which can be a number, a letter, a symbol, etc. after vector conversion by a word vector model. Each main business type corresponds to a unique enterprise cluster number. It needs to be further explained that the first representation vector includes vector data of the main business of the target enterprise and vector data of the first enterprise cluster number, for example, vector [O, p], wherein O is the vector data corresponding to the first enterprise cluster number, and p is the vector data of the main business.

[0056] S103, match the first enterprise clustering number with a preset enterprise clustering library to obtain a clustering cluster corresponding to the first enterprise clustering number, the preset enterprise clustering library including a corresponding relationship between clustering numbers and clustering clusters, and the clustering cluster corresponding to the first enterprise clustering number including main business documents of a plurality of enterprises having a same main business type as the target enterprise.

[0057] In the above step, the first enterprise clustering number represents the main business type of the target enterprise, and the first enterprise clustering number is matched with the preset enterprise clustering library according to the corresponding relationship between the clustering numbers and the clustering clusters stored in the preset enterprise clustering library to obtain the clustering cluster corresponding to the main business type of the target enterprise, and at this time, the clustering cluster corresponding to the main business type of the target enterprise includes main business documents of a plurality of enterprises having a same main business type as the target enterprise.

[0058] S104, obtaining a plurality of enterprises having a same main business type as the target enterprise based on the clustering cluster corresponding to the first enterprise clustering number.

[0059] In the above step, the target enterprise's main business is matched with the clustering cluster corresponding to the first enterprise clustering number to obtain a plurality of enterprises having a same main business type as the target enterprise, and the enterprise information of the plurality of enterprises is output as an output result and displayed on the user equipment of the user. The clustering cluster can be understood as a data set that aggregates the same main business type information. For example, if the main business type of a clustering cluster is wheat, the clustering cluster includes hard wheat, soft wheat and other wheat. Because the number of main businesses is large, the clustering cluster classifies and summarizes the large main business data, aggregates the main business data with high similarity to the same clustering cluster, and divides the main business data with low similarity to different clustering clusters, thereby reducing the data processing amount during data matching. The specific process of constructing the enterprise clustering library is as follows: first, obtain a main business classification tree, the main business classification tree storing detailed data of all main business types and enterprise clustering numbers of each main business type, and dividing the main business type into a preset number of product classification levels, the product classification level reflecting the product range of the main business type, and the application embodiment is preferably 5 product classification levels, for example, the 1st product classification can be agricultural products, the 2nd product classification can be grains, the 3rd product classification can be rice, the 4th product classification can be stem rice, and the 5th product classification can be stem rice for seed. Of course, the number of product classification levels is determined according to the actual situation, and is not limited here.

[0060] Then, the plurality of main businesses in the preset vector representation database are clustered by the plurality of enterprise clustering numbers stored in the main business classification tree to obtain a plurality of clustering clusters, wherein one enterprise clustering number corresponds to one clustering cluster. The specific process is as follows:

[0061] Taking the clustering process of the second main business as an example, the second main business is any one of multiple main businesses in the database represented by a preset vector. Then, the center vectors of the representation vector of the second main business and the cluster numbers of multiple enterprises are obtained. The center vector of the enterprise cluster number can be understood as the cluster center of the cluster. Within a preset range of the cluster center, all data belong to the same cluster. The selection of the cluster center can be based on the product classification of the main business type. This application preferably uses the vector of the main business type with a 3-level product classification as the center vector of the cluster center. It should be explained that the representation vector of the second main business and the center vector of the enterprise cluster number are vectors in the same vector space. Then, the cosine distance between the center vectors of the second main business and the cluster numbers of multiple enterprises is calculated. The cosine distance reflects the similarity between the main business type of the second main business and the cluster.

[0062] In one possible implementation, if the cosine distance between the representation vector of the second main business and the center vector of the second enterprise cluster number is greater than or equal to a preset threshold, then the representation vector of the second main business and the second enterprise cluster number are established as a correspondence, that is, the second main business is assigned to the cluster corresponding to the second enterprise cluster number, wherein the second enterprise cluster number is any one of multiple enterprise cluster numbers.

[0063] In another possible implementation, if the first cosine distance between the representation vector of the second main business and the center vector of the second enterprise cluster number is equal to the second cosine distance between the representation vector of the second main business and the center vector of the third enterprise cluster number, and both the first cosine distance and the second cosine distance are greater than or equal to a preset threshold, then the second main business is simultaneously associated with both the second and third cluster numbers to ensure the data integrity of each cluster. The third enterprise cluster number is any one of multiple enterprise cluster numbers.

[0064] Finally, the cluster numbers of multiple enterprises are associated with their respective clusters and stored in a pre-defined enterprise cluster library.

[0065] This application also provides a rapid retrieval system for searching similar companies, the system being a server, such as... Figure 3 As shown, the server includes an acquisition module 1 and a processing module 2, wherein:

[0066] Module 1 is used to obtain the main business of the target company as input by the user;

[0067] The processing module 2 is configured to match the main business with a preset vector representation database to obtain a first representation vector, the first representation vector comprising a first enterprise cluster number, the first enterprise cluster number representing a vector number of a main business type of the target enterprise, and the preset vector representation database comprising a corresponding relationship between the main business and a corresponding representation vector; match the first enterprise cluster number with a preset enterprise cluster library to obtain a cluster cluster corresponding to the first enterprise cluster number, the preset enterprise cluster library comprising a corresponding relationship between a cluster number and a cluster cluster, and the cluster cluster corresponding to the first enterprise cluster number comprising main business documents of a plurality of enterprises having the same main business type as the target enterprise; and obtain a plurality of enterprises having the same main business type as the target enterprise based on the cluster cluster corresponding to the first enterprise cluster number.

[0068] In a possible implementation, before matching the main business with the preset vector representation database to obtain the first representation vector, the method comprises:

[0069] The acquisition module 1 acquires a plurality of enterprise introduction texts; extracts main businesses in the plurality of introduction texts by using a preset rule screening model, the preset rule screening model being defined with a plurality of information screening rules, wherein one introduction text corresponds to one main business, and one main business comprises a plurality of main business words;

[0070] The processing module 2 inputs the first main business into a plurality of preset pre-training models to obtain a plurality of first text vectors, the first main business being any one of the plurality of main businesses; performs first pooling processing on the plurality of first text vectors respectively to generate a plurality of second text vectors; performs second pooling processing on the plurality of second text vectors to obtain a third text vector; and stores the third text vector as a representation vector corresponding to the first main business in the preset vector representation database.

[0071] In a possible implementation, before matching the main business with the preset vector representation database to obtain the first representation vector, the method further comprises: the acquisition module 1 retrieves a plurality of word vector models from a preset word vector model library; constructs the main businesses of the plurality of enterprises into a training sample library; and the processing module 2 inputs data in the training sample library into the plurality of word vector models to obtain a plurality of training text vectors; calculates an evaluation value corresponding to each of the plurality of word vector models by using a vector space distribution evaluation function based on the plurality of training text vectors; sorts the evaluation values corresponding to each of the plurality of word vector models from large to small to generate a sorting table; and selects a plurality of word vector models having a ranking greater than or equal to a preset ranking in the sorting table as pre-training models.

[0072] In a possible implementation, based on the plurality of training text vectors, the evaluation value corresponding to each of the plurality of word vector models is calculated by using a vector space distribution evaluation function, specifically:

[0073]

[0074] wherein T is the evaluation value, W xi is the training text vector corresponding to the i-th main business word of the x-th main business, W yj is the training text vector corresponding to the i-th main business word of the y-th main business, n is the number of main business categories, m is the number of main business words in the n-th main business category, the numerator is used to represent the average similarity between the plurality of main business words of the same main business, and the denominator is used to represent the average similarity value between the plurality of main business words corresponding to different main businesses.

[0075] In a possible implementation, the processing module 2 traverses the word vector dimensions of the plurality of word vector models; if the first word vector dimension is less than or equal to the second word vector dimension, the first word vector dimension is determined as the standard vector dimension, and the first word vector dimension and the second word vector dimension are any two word vector dimensions in the plurality of word vector dimensions; and the word vector dimensions corresponding to the plurality of word vector models are respectively compressed to the standard vector dimension.

[0076] In a possible implementation, before the first enterprise cluster number is matched with the preset enterprise cluster library to obtain the cluster cluster corresponding to the first enterprise cluster number, the method further includes: the acquisition module 1 determines a plurality of enterprise cluster numbers of the preset enterprise cluster library based on the main business classification tree; the processing module 2 clusters the representation vectors of the plurality of main businesses in the preset vector representation database based on the plurality of enterprise cluster numbers to obtain a plurality of cluster clusters, wherein one enterprise cluster number corresponds to one cluster cluster; and the plurality of enterprise cluster numbers and the cluster clusters corresponding to the plurality of enterprise cluster numbers are constructed into a corresponding relationship and stored in the preset enterprise cluster library.

[0077] In a possible implementation, the acquisition module 1 acquires the representation vector of the second main business and the center vector of the plurality of enterprise cluster numbers, and the second main business is any one of the plurality of main businesses in the preset vector representation database; when the cosine distance value between the representation vector of the second main business and the center vector of the second enterprise cluster number is greater than or equal to a preset threshold, the processing module 2 constructs the representation vector of the second main business and the second enterprise cluster number into a corresponding relationship, and the second enterprise cluster number is any one of the plurality of enterprise cluster numbers.

[0078] It should be noted that the apparatus provided in the above examples is only used as an example for the division of the above functional modules in realizing its functions, and in actual application, the above functions can be completed by different functional modules according to needs, that is, the internal structure of the device is divided into different functional modules to complete all or part of the above described functions. In addition, the apparatus and method embodiments provided in the above examples belong to the same concept, and the specific implementation process is detailed in the method embodiments, which will not be described here.

[0079] The present application also provides an electronic device. Referring to Figure 4 , Figure 4 is a structural schematic diagram of an electronic device disclosed in an embodiment of the present application. The electronic device 400 can include at least one processor 401, at least one network interface 404, a user interface 403, a memory 405, and at least one communication bus 402.

[0080] The communication bus 402 is used to realize the connection and communication between the components.

[0081] The user interface 403 can include a display screen (Display), a camera (Camera), and optionally a standard wired interface and a wireless interface.

[0082] The network interface 404 can optionally include a standard wired interface and a wireless interface (such as a WI-FI interface).

[0083] The processor 401 can include one or more processing cores. The processor 401 connects various parts within the server through various interfaces and lines, performs various functions of the server and processes data by running or executing instructions, programs, code sets or instruction sets stored in the memory 405, and calling data stored in the memory 405. Alternatively, the processor 401 can be implemented in at least one of a hardware form of a digital signal processing (DSP), a field-programmable gate array (FPGA), and a programmable logic array (PLA). The processor 401 can integrate a combination of one or more of a central processing unit (CPU), a graphics processing unit (GPU), and a modem. Among them, the CPU mainly processes operating systems, user interfaces, and application programs; the GPU is responsible for rendering and drawing the content to be displayed on the display screen; and the modem is used for processing wireless communication. It can be understood that the above-mentioned modem can also not be integrated into the processor 401, but can be realized by a separate chip.

[0084] The memory 405 can include a random access memory (RAM) and can also include a read-only memory (ROM). Optionally, the memory 405 includes a non-transitory computer-readable storage medium. The memory 405 can be used to store instructions, programs, codes, code sets or instruction sets. The memory 405 can include a program storage area and a data storage area, wherein the program storage area can store instructions for implementing an operating system, instructions for at least one function (such as a touch function, a sound playing function, an image playing function, etc.), instructions for implementing the above-mentioned various method embodiments, etc.; the data storage area can store data involved in the above-mentioned various method embodiments, etc. The memory 405 can also be at least one storage device located away from the aforementioned processor 401. Referring to Figure 4 The memory 405 as a computer storage medium can include an operating system, a network communication module, a user interface module, and an application program of a fast retrieval method for querying similar enterprises.

[0085] In Figure 4In the electronic device 400 shown, the user interface 403 is mainly used to provide an interface for the user to input, and obtain data input by the user; and the processor 401 can be used to invoke an application program stored in the memory 405 and storing a fast retrieval method for querying similar enterprises, which, when executed by one or more processors 401, causes the electronic device 400 to perform the method described in one or more of the above embodiments. It should be noted that, for the above-mentioned method embodiments, in order to simply describe, they are all described as a combination of a series of actions, but those skilled in the art should know that the present application is not limited to the action sequence described, because according to the present application, certain steps can be performed in other sequences or at the same time. Secondly, those skilled in the art should know that the embodiments described in the specification all belong to preferred embodiments, and the actions and modules involved are not necessarily required by the present application.

[0086] In the above embodiments, the description of each embodiment has its own focus, and the parts not described in detail in a certain embodiment can be referred to the related description of other embodiments.

[0087] In several embodiments provided by the present application, it should be understood that the disclosed device can be implemented in other ways. For example, the device embodiments described above are only schematic. The division of the units is only a logical function division. There can be another division manner in actual implementation. For example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the displayed or discussed mutual couplings or direct couplings or communication connections between different units, can be indirect couplings or communication connections through some interfaces, devices or units, and can be electrical or other forms.

[0088] The units described as separate components can or can not be physically separate, and the components shown as units can or can not be physical units, i.e. can be located in one place, or can be distributed on a plurality of network units. Some or all of the units can be selected according to actual needs to achieve the purpose of the embodiment.

[0089] In addition, each functional unit in each embodiment of the present application can be integrated into a processing unit, or each unit can exist physically independently, or two or more units can be integrated into one unit. The integrated unit can be realized in the form of hardware or in the form of a software functional unit.

[0090] The integrated unit, if implemented in the form of a software function unit and sold or used as an independent product, can be stored in a computer readable memory. Based on such understanding, the technical solutions of the present application essentially or the part that contributes to the prior art or the whole or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a memory and includes a plurality of instructions for causing a computer device (which can be a personal computer, a server or a network device, etc.) to execute all or part of the steps of the embodiments of the present application. The aforementioned memory includes: a U disk, a mobile hard disk, a magnetic disk or an optical disk, and various media that can store program codes.

[0091] The above-described are only exemplary embodiments of the present disclosure, and cannot limit the scope of the present disclosure. That is, any equivalent changes and modifications made in accordance with the teachings of the present disclosure are still within the scope of the present disclosure. Other embodiments of the present disclosure will be readily apparent to those skilled in the art upon considering the specification and practicing the true principles of the present disclosure.

[0092] The present application is intended to cover any variations, uses, or adaptive changes of the present disclosure that follow the general principles of the present disclosure and include common knowledge or conventional technical means in the technical field not recorded in the present disclosure. The specification and examples are only considered as exemplary, and the scope and spirit of the present disclosure are defined by the claims.

Claims

1. A quick search method for querying similar enterprises, characterized in that, Applied to a server, the method comprises: obtaining the main business of the target enterprise input by the user; matching the main business with a preset vector representation database to obtain a first representation vector, the first representation vector comprising a first enterprise cluster number, the first enterprise cluster number representing a vector number of the main business type of the target enterprise, the preset vector representation database comprising a correspondence between main businesses and their corresponding representation vectors; matching the first enterprise cluster number with a preset enterprise cluster library to obtain a cluster cluster corresponding to the first enterprise cluster number, the preset enterprise cluster library comprising a correspondence between cluster numbers and cluster clusters, and the cluster cluster corresponding to the first enterprise cluster number comprising main business documents of a plurality of enterprises with the same main business type as the target enterprise; based on the cluster cluster corresponding to the first enterprise cluster number, obtaining a plurality of enterprises with the same main business type as the target enterprise; before matching the main business with the preset vector representation database to obtain the first representation vector, comprising: obtaining a plurality of introduction texts of enterprises; using a preset rule screening model to extract the main business from the plurality of introduction texts, the preset rule screening model defining a plurality of information screening rules, wherein one introduction text corresponds to one main business, and one main business comprises a plurality of main business words; inputting the first main business into a plurality of pre-training models to obtain a plurality of first text vectors, the first main business being any one of the plurality of main businesses; performing first pooling processing on each of the plurality of first text vectors to generate a plurality of second text vectors; performing second pooling processing on the plurality of second text vectors to obtain a third text vector; storing the third text vector as the representation vector corresponding to the first main business in the preset vector representation database; before matching the main business with the preset vector representation database to obtain the first representation vector, further comprising: calling a plurality of word vector models from a preset word vector model library; constructing the main businesses of the plurality of enterprises as a training sample library; inputting the data in the training sample library into the plurality of word vector models to obtain a plurality of training text vectors; based on the plurality of training text vectors, using a vector space distribution evaluation function to calculate an evaluation value corresponding to each of the plurality of word vector models; sorting the evaluation values corresponding to each of the plurality of word vector models from large to small to generate a ranking table; ranking the plurality of word vector models with a ranking greater than or equal to a preset ranking as pre-training models; the method of calculating an evaluation value corresponding to each of the plurality of word vector models based on the plurality of training text vectors using a vector space distribution evaluation function, specifically comprising: Wherein, T is an evaluation value, Wxi is the training text vector corresponding to the i th main business word of the x th main business, Wyj is the training text vector corresponding to the i th main business word of the y th main business, n is the number of main business categories, m is the number of main business words in the n th main business category, the numerator is used to represent the average similarity between the multiple main business words of the same main business, and the denominator is used to represent the average similarity value between the multiple main business words corresponding to different main businesses.

2. The method of claim 1, wherein, The method further comprises: traversing the word vector dimensions of the plurality of word vector models; if the first word vector dimension is less than or equal to the second word vector dimension, determining the first word vector dimension as the standard vector dimension, the first word vector dimension and the second word vector dimension being any two word vector dimensions in the plurality of word vector dimensions; compressing the word vector dimensions corresponding to the plurality of word vector models to the standard vector dimension respectively.

3. The method of claim 1, wherein, Before matching the first enterprise cluster number with the preset enterprise cluster library to obtain the cluster cluster corresponding to the first enterprise cluster number, the method further comprises: determining a plurality of enterprise cluster numbers of the preset enterprise cluster library based on the main business classification tree; based on the plurality of enterprise cluster numbers, clustering the representation vectors of the plurality of main businesses in the preset vector representation database to obtain a plurality of cluster clusters, wherein one enterprise cluster number corresponds to one cluster cluster; 4. The method of claim 3, wherein, constructing a corresponding relationship between the plurality of enterprise cluster numbers and the cluster clusters corresponding to the plurality of enterprise cluster numbers, and storing the corresponding relationship in the preset enterprise cluster library. The clustering of the representation vectors of the plurality of main businesses in the preset vector representation database based on the plurality of enterprise cluster numbers comprises:

5. A rapid retrieval system for finding similar companies, characterized in that, obtaining the representation vector of a second main business and the center vector of a plurality of enterprise cluster numbers, the second main business being any one of the plurality of main businesses in the preset vector representation database; when the cosine distance value between the representation vector of the second main business and the center vector of a second enterprise cluster number is greater than or equal to a preset threshold, constructing a corresponding relationship between the representation vector of the second main business and the second enterprise cluster number, the second enterprise cluster number being any one of the plurality of enterprise cluster numbers. The system for performing the fast retrieval method for querying similar enterprises according to any one of claims 1 to 4, the system is a server, the server comprises an acquisition module (1) and a processing module (2), wherein: the acquisition module (1) is used for acquiring the main business of the target enterprise input by the user; The processing module (2) is configured to match the main business with a preset vector representation database to obtain a first representation vector, the first representation vector comprises a first enterprise cluster number, the first enterprise cluster number represents a vector number of a main business type of the target enterprise, the preset vector representation database comprises a corresponding relationship between a main business and a corresponding representation vector; match the first enterprise cluster number with a preset enterprise cluster library to obtain a cluster cluster corresponding to the first enterprise cluster number, the preset enterprise cluster library comprises a corresponding relationship between a cluster number and a cluster cluster, the cluster cluster corresponding to the first enterprise cluster number comprises main business documents of a plurality of enterprises with the same main business type as the target enterprise; and obtain a plurality of enterprises with the same main business type as the target enterprise based on the cluster cluster corresponding to the first enterprise cluster number.

6. An electronic device, comprising: The electronic device (400) comprises a processor (401), a memory (405), a user interface (403) and a network interface (404), the memory (405) is configured to store instructions, the user interface (403) and the network interface (404) are configured to communicate with other devices, and the processor (401) is configured to execute the instructions stored in the memory (405) to enable the electronic device (400) to perform the method of any one of claims 1 to 4.

7. A computer-readable storage medium, characterized in that, The computer readable storage medium stores instructions, when the instructions are executed, the method of any one of claims 1 to 4 is executed.

Images