Enterprise actual operation address identification method and device, computer equipment and storage medium

Abstract

The invention discloses an enterprise actual operation address identification method and device, computer equipment and a storage medium. The method comprises the following steps: collecting related data from different data sources, and extracting address information to obtain a plurality of extraction results; fusing the plurality of extraction results to obtain final address information; and outputting the final address information. By implementing the method provided by the invention, the problems of incomplete data coverage, disordered information, limited accuracy and difficulty in timely updating when the actual operation address of an enterprise is identified by RPA, NLP and OCR in the prior art can be solved.

Description

Technical Field

[0001] The present invention relates to an address recognition method, and more specifically to a method, device, computer equipment and storage medium for recognizing the actual business address of an enterprise. Background Art

[0002] In the modern business environment, accurate identification of the actual business address of an enterprise is crucial for market analysis, customer relationship management, and compliance inspection. Currently, the industry mainly relies on artificial intelligence technologies such as RPA (Robotic Process Automation), address standardization and cleaning, NLP (Natural Language Processing), and OCR (Optical Character Recognition) to solve this problem. Although these technologies have made some progress, they still face many challenges and limitations in the application process.

[0003] It is common to use RPA technology to obtain corporate address information from recruitment websites, corporate yellow pages, self-built websites, and online information bases of the Industrial and Commercial Bureau. However, this method has limited data coverage, especially for companies that have not disclosed their detailed business information on the Internet, it is difficult to obtain comprehensive information. In addition, the data on the Internet is often disorganized, mixed with a lot of noise and invalid information, and requires a lot of manpower to clean and filter. For market competition strategy considerations, some companies choose not to fully disclose their business details, which makes it difficult to obtain accurate main product or service information relying solely on public channels. Therefore, technical means based on existing public data have certain limitations in accuracy. In order to solve the problems of address expression diversity and authenticity verification, address standardization and cleaning technology uses a large-scale address corpus combined with NLP algorithms to process raw data. However, to ensure that the final output address data is both accurate and consistent, it usually requires a large number of high-quality training samples to support it, and with the emergence of new place names and social changes, the model also needs to be continuously updated to maintain optimal performance. NLP technology can help companies extract potential business address clues from text resources such as press releases, product introductions, and research reports. However, due to the lack of authorized access rights, many key internal documents cannot be included in the analysis scope, limiting the application effect of this method. At the same time, the ambiguity and context-dependent characteristics of natural language itself also bring additional difficulties to accurate recognition. OCR technology is good at automatically reading and converting address content from paper documents or scans into electronic text format. However, its recognition accuracy is easily affected by factors such as image quality, font style and typesetting rules, and sometimes manual review is required to ensure the correctness of the results. Over time, urban planning adjustments, street naming changes, and new properties will cause changes in address information. Existing static databases are difficult to reflect these latest changes in a timely manner, which affects subsequent business decisions based on addresses.

[0004] In summary, although there are currently a variety of technologies applied to the identification of the actual business addresses of enterprises, they still need to be further optimized and developed in terms of the integrity of the data source, the amount of data required for model training, the ability to track the evolution of address information over time, and the effective processing of unstructured data.

[0005] Therefore, it is necessary to design a new method to solve the problems faced by existing technologies such as RPA, NLP and OCR when identifying the actual business address of an enterprise, such as incomplete data coverage, messy information, limited accuracy and difficulty in timely updating. Summary of the invention

[0006] The purpose of the present invention is to overcome the defects of the prior art and provide a method, device, computer equipment and storage medium for identifying the actual business address of an enterprise.

[0007] To achieve the above purpose, the present invention adopts the following technical solution: a method for identifying the actual business address of an enterprise, comprising:

[0008] Collect relevant data from different data sources and extract address information to obtain multiple extraction results;

[0009] Merging the multiple extraction results to obtain final address information;

[0010] The final address information is output.

[0011] Its further technical solution is: the data sources include corporate official websites, recruitment platforms, corporate information query platforms, social media and public accounts.

[0012] A further technical solution is: collecting relevant data from different data sources and extracting address information to obtain multiple extraction results, including:

[0013] When collecting addresses from the official website of an enterprise, obtain the enterprise name, search for official links based on the enterprise name, and locate and extract text information containing keywords from the website corresponding to the official link, and use the text information as the extraction result;

[0014] When collecting addresses from the recruitment platform, obtain the company name, search for relevant recruitment information based on the company name, and locate and extract text information containing keywords from the relevant recruitment information, and use the text information as the extraction result;

[0015] When collecting addresses from the enterprise information query platform, obtaining the enterprise name or unified social credit code, searching for relevant enterprise information based on the enterprise name or unified social credit code, and locating and extracting text information containing keywords from the relevant enterprise information, and using the text information as the extraction result;

[0016] When collecting addresses from social media and official accounts, search for the official social media and official accounts of the enterprise, and extract text information containing keywords, and use the text information as the extraction result.

[0017] A further technical solution is: the fusion of the multiple extraction results to obtain the final address information includes:

[0018] When all extraction results are inconsistent, a voting mechanism is used to determine the final address information; when there are multiple identical addresses in multiple extraction results, a voting mechanism is used to select the address with the most occurrences as the final address information.

[0019] A further technical solution is: the fusion of the multiple extraction results to obtain the final address information includes:

[0020] When the high-level information in multiple extraction results is the same but the low-level information is different, the high-level information is retained and the low-level information is merged to form the final address information, wherein the high-level information includes the street name and the building name, and the low-level information includes the floor number.

[0021] A further technical solution is: collecting relevant data from different data sources and extracting address information to obtain multiple extraction results, including:

[0022] When some of the multiple extraction results are ambiguous, the details supplement and consistency priority rules are used to determine the final address information.

[0023] A further technical solution is: the fusion of the multiple extraction results to obtain the final address information also includes:

[0024] When the quantities in multiple extraction results are the same, the final address information is determined according to the preset priority rules.

[0025] The present invention also provides a device for identifying the actual business address of an enterprise, comprising:

[0026] An extraction unit, used to collect relevant data from different data sources and extract address information to obtain multiple extraction results;

[0027] A fusion unit, used for fusing the multiple extraction results to obtain final address information;

[0028] An output unit is used to output the final address information.

[0029] The present invention further provides a computer device, which includes a memory and a processor. The memory stores a computer program, and the processor implements the above method when executing the computer program.

[0030] The present invention also provides a storage medium, wherein the storage medium stores a computer program, and the computer program implements the above method when executed by a processor.

[0031] Compared with the prior art, the present invention has the following beneficial effects: the present invention collects relevant data from multiple data sources to ensure comprehensive information coverage, thus avoiding the problem of incomplete data coverage in traditional technologies; then, by combining the language model module in the intelligent body, multiple candidate address information is extracted, thus solving the problem of messy information in traditional technologies. Using data fusion technology, multiple extraction results are integrated to improve the accuracy of address information, thus overcoming the defect of limited accuracy of existing methods. Through intelligent algorithm optimization, the system can update and correct address information in real time, thus solving the problem that traditional technologies cannot be updated in a timely manner. Finally, accurate, comprehensive and up-to-date final address information is output, which greatly improves the efficiency and reliability of business address identification.

[0032] The present invention is further described below in conjunction with the accompanying drawings and specific embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying any creative work.

[0034] Figure 1 A schematic diagram of an application scenario of the method for identifying the actual business address of an enterprise provided by an embodiment of the present invention;

[0035] Figure 2 A schematic diagram of a process for identifying the actual business address of an enterprise provided by an embodiment of the present invention;

[0036] Figure 3 A schematic block diagram of a device for identifying the actual business address of an enterprise provided by an embodiment of the present invention;

[0037] Figure 4 A schematic block diagram of a computer device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0038] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0039] It should be understood that when used in this specification and the appended claims, the terms "include" and "comprises" indicate the presence of described features, integers, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integers, steps, operations, elements, components and / or combinations thereof.

[0040] It should also be understood that the terms used in this specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. As used in the specification of the present invention and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include plural forms.

[0041] It should be further understood that the term "and / or" used in the present description and the appended claims refers to any and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0042] See also Figure 1 and Figure 2 , Figure 1 A schematic diagram of an application scenario of the method for identifying the actual business address of an enterprise provided in an embodiment of the present invention. Figure 2 A schematic flow chart of a method for identifying the actual business address of an enterprise provided in an embodiment of the present invention. The method for identifying the actual business address of an enterprise is applied in a server. The server interacts with the terminal to collect relevant data from different data sources through different agents, and extracts address information to obtain multiple extraction results; fuses the multiple extraction results to obtain final address information; and outputs the final address information. This solves the problems of incomplete data coverage, messy information, limited accuracy and difficulty in timely updating faced by existing technologies such as RPA, NLP and OCR when identifying the actual business address of an enterprise.

[0043] Figure 2 FIG. 1 is a flow chart of a method for identifying the actual business address of an enterprise provided by an embodiment of the present invention. Figure 2 As shown, the method includes the following steps S110 to S130.

[0044] S110 , collecting relevant data from different data sources and extracting address information to obtain multiple extraction results.

[0045] In this embodiment, the data sources include corporate official websites, recruitment platforms, corporate information query platforms, social media and public accounts. The extraction results refer to the address information collected from different data sources.

[0046] Specifically, relevant data is collected from different data sources, and address information is extracted. When multiple identical addresses are collected during the extraction of address information, a voting mechanism is used to select the address with the most occurrences as the extraction result.

[0047] In addition, relevant data is collected from different data sources, and address information is extracted. When the amount of newly collected address information does not meet the requirements when extracting the address information, the collected address information is retained as the extraction result.

[0048] In one embodiment, the above step S110 may include steps S111 to S114.

[0049] S111. When collecting addresses from the official website of an enterprise, obtain the enterprise name, search for official links according to the enterprise name, and locate and extract text information containing keywords from the website corresponding to the official link, and use the text information as the extraction result.

[0050] In this embodiment, agent 1 is used to obtain the real address from the enterprise official website. Agent 1 includes: a request module, a search module, a language model module and an output module.

[0051] Request module: for users to enter the company name.

[0052] Search module: Based on the company name entered, the system searches the Internet for the three most relevant company official website links.

[0053] The language model module uses RPA (Robotic Process Automation), NLP (Natural Language Processing), OCR (Optical Character Recognition), etc.: it provides specific prompt words to guide the system to automatically access the provided links and locate modules on the website such as "Company Introduction", "About Us" or "Company Address" that may contain address information.

[0054] Specifically, the prompt words:

[0055] Website access: Open the provided recruitment platform link {url} and enter the official website of the target company.

[0056] Information module positioning: Find the following modules on the website: "Business Information", "Work Address", "Company Introduction". These modules may contain the detailed address information we need.

[0057] Address information identification: Read the contents of the above modules carefully and look for detailed address information such as "No. 7, E Road, D Park, District C, Province A, City B". Make sure the identified address information contains the province, city, region, street name and specific house number.

[0058] Information extraction and recording: Once the address information that meets the format requirements is found, the corresponding information is accurately recorded and provided. If the same address information is found in multiple modules, a voting mechanism can be used to select the address with the most occurrences as the final result.

[0059] Special case handling: If the address information found in different modules is inconsistent, determine which address is more reliable based on the level of detail and completeness of the information. If the address information is missing or incomplete, ignore it and continue searching in other modules.

[0060] The system will carefully read the contents of these modules, looking for detailed address information that conforms to a specific format (province, city, region, street name and specific house number).

[0061] If the same address information is found in multiple modules, a voting mechanism is used to select the address with the most occurrences as the final result. If the address information is inconsistent, the more reliable one is determined based on the level of detail and completeness; if the information is missing or incomplete, it is ignored and other modules are searched.

[0062] Output module: finally outputs the address information extracted from the three most relevant corporate official websites.

[0063] S112. When collecting addresses from the recruitment platform, obtain the company name, search for relevant recruitment information according to the company name, and locate and extract text information containing keywords from the relevant recruitment information, and use the text information as the extraction result.

[0064] In this embodiment, the agent 2 used in this step is similar to agent 1. Enter the company name and enter the request module. In the search module, find the recruitment platform, positions and related links related to the input company name. Enter the LLM module, find the "business information", "work address", "company introduction" and other modules in the recruitment platform, and extract detailed address information. According to the identified address information, record and output the company address in the recruitment platform.

[0065] Recruitment platforms are frequently updated and may contain the latest company addresses or work locations, which are particularly useful for quickly obtaining the company's office location during recruitment activities. The address information on recruitment platforms can be improved by comparing the content of different modules (such as company introduction and work address). These platforms are directly aimed at job seekers, and the addresses provided usually represent the specific locations of the company's actual operations and recruitment needs, which are highly practical.

[0066] S113. When collecting addresses from the enterprise information query platform, obtain the enterprise name or unified social credit code, search for relevant enterprise information based on the enterprise name or unified social credit code, and locate and extract text information containing keywords from the relevant enterprise information, and use the text information as the extraction result.

[0067] In this embodiment, the third agent is similar to the previous two agents.

[0068] Enter the company name and enter the request module. In the search module, find three company links related to the entered company name, pointing to the enterprise information query platform. Search for company details on the platform by company name or unified social credit code. Enter the LLM module, search for "basic information", "business information", "address" and other modules on the platform, and extract address information. Output and record the address information obtained from the enterprise information query platform.

[0069] Corporate information query platforms usually obtain data from official databases such as industrial and commercial registration, ensuring the legitimacy and accuracy of the information. The platform may cover more company information, such as historical addresses, addresses of different branches, etc., to facilitate finding the most accurate company location. Since it is an official source, the address provided by the query platform can generally be used as a reference for other data to ensure the consistency and accuracy of information from various channels.

[0070] S114. When collecting addresses from social media and public accounts, search for the official social media and public accounts of the enterprise, and extract text information containing keywords, and use the text information as the extraction result.

[0071] In this embodiment, Agent 4 is similar to the previous three agents. It inputs the company name and enters the request module. In the search module, it finds the social media accounts and public accounts related to the input company name. It enters the LLM module (language model module) and searches for address information in the "Company Profile" or "Contact Us" modules of the social media account or public account. The output module records and outputs the company address in the social media and public account based on the identified address information.

[0072] Social media and public account information are usually more interactive and updated, reflecting the company's recent activities and address changes. Many companies publish company information on social platforms, including office addresses, contact information, etc., so address information can be confirmed through multiple platforms. Through social platforms, users (such as customers and employees) may mention the company's address in comments, and the company's address information can be indirectly verified through this information.

[0073] The address information of the same enterprise is collected through four different channels (the official website of the enterprise, the recruitment platform, the enterprise information query platform, and social media and public accounts), which increases the diversity and accuracy of the data. This method helps to reduce the information bias or errors that may be caused by a single source. Each agent is designed with clear information recognition and processing rules, including specific requirements for the address information format, which greatly improves the accuracy and reliability of the address information obtained. When the address information provided by different sources is inconsistent, the system not only relies on the simple majority rule, but also considers the detail and completeness of the information to make more reasonable decisions. The whole process is completed automatically by the system, reducing the need for manual intervention and improving work efficiency. At the same time, for the processing of a large amount of enterprise information, such an automated process can significantly save time and resources. This method is suitable for different types of enterprises, regardless of their size or industry field, as long as these enterprises have corresponding public information on the Internet, they can be effectively processed. It provides users with a fast and convenient way to obtain the real address information of the enterprises they are interested in, improving the user experience and service quality.

[0074] By obtaining the same address information through different platforms and modules, the most credible results can be selected through a voting mechanism, thereby improving accuracy. The information provided by different modules may have different focuses. The official address provided by the official website may complement the addresses obtained from recruitment platforms, social media, etc. to avoid omissions. Recruitment platforms and social media information are updated frequently, while the addresses of corporate information query platforms and official websites are stable. The combination of the two can obtain address data that is both accurate and timely. From official websites to social media, and then to recruitment platforms, multiple sources can provide information in different scenarios to ensure that the address information obtained is not missed.

[0075] In summary, by integrating data from different online resources and combining intelligent algorithms and rule settings, this solution aims to provide an efficient, accurate and reliable solution for obtaining corporate address information.

[0076] S120: Merge the multiple extraction results to obtain final address information.

[0077] In this embodiment, the final address information refers to the finally determined address information.

[0078] Specifically, when all extraction results are inconsistent, a voting mechanism is used to determine the final address information; when there are multiple identical addresses in multiple extraction results, a voting mechanism is used to select the address with the most appearances as the final address information. Let f(a) be the number of times address a appears in set B, then: the final address is = argmax a∈B f(a).

[0079] When the high-level information in multiple extraction results is the same but the low-level information is different, the high-level information is retained and the low-level information is merged to form the final address information, wherein the high-level information includes the street name and the building name, and the low-level information includes the floor number.

[0080] For example: Agent 1 and Agent 2 provide "5th floor, Building A, Center D, District C, Province A, City B";

[0081] Agent 3 and Agent 4 provided “6th Floor, Building A, Center, District C, Province A, City B”.

[0082] The final output is: "5-6th Floor, Building A, D Center, District C, City B, Province A".

[0083] When some of the multiple extraction results are ambiguous, the details supplement and consistency priority rules are used to determine the final address information.

[0084] For example: Agent 1 provides “A Province B City C District D Center A Building;

[0085] Agents 2 to 4 provided “A Province B City C District D Center A Floor 6”.

[0086] The final output is: "6th Floor, Building A, D Center, District C, City B, Province A".

[0087] When the quantities in multiple extraction results are the same, the final address information is determined according to the preset priority rules.

[0088] In this embodiment, when the collected address information, that is, the address information of the partial extraction result, is insufficient, the system will generate the final address set based on the acquired data. Assuming that the expected address set is A, and the actual collected address is B, the final output address set will be B. This mechanism ensures that even if the data is incomplete, the system can still give the most likely answer.

[0089] Even if some intelligent agents do not collect data, the system can output a reasonable result based on the existing data, reducing decision-making failures caused by missing data.

[0090] If multiple address information appears repeatedly, the system uses a voting mechanism to select the address with the most occurrences as the final result. For example, if Agents 1 and 2 both provide the address "5th Floor, Building A, D Center, District C, Province A, City B", and the addresses from other sources are not exactly the same, the system will select the most frequently appearing address by calculating the number of occurrences of each address in the collection results.

[0091] For example: Agents 1 and 2 provide "5th floor, Building A, Center D, District C, Province A, City B";

[0092] Agent 3 provides "A Province B City C District D Center A Building 6th Floor";

[0093] Intelligent agent 4 provided “7th floor, Building A, Center, District C, Province A, City B”.

[0094] The final output is: "5th Floor, Building A, D Center, District C, City B, Province A".

[0095] The voting mechanism can effectively reduce the impact of accidental errors or abnormal data on the final results, and ensure that the consensus of the majority of data sources can dominate the final decision.

[0096] When the voting results are the same or the address information is similar, priority rules can be designed, such as giving priority to sources with higher reliability or more detailed information. If there are two groups of addresses with the same number, a priority function P(a) can be defined, where a is the address, then:

[0097] For example, in the collected address information, if the addresses of agents one and two are the same, but agent two provides more floor information, the address of agent two can be selected preferentially.

[0098] Priority rules can accurately filter out the most credible addresses based on the quality, reliability and detail of the data, avoiding simple duplicate addresses that interfere with decision-making.

[0099] In the event of an address conflict, the system will choose the address with more detailed information. For example, if one address only gives the name of the building, and another address provides detailed floor information, the system may give priority to the address with more details. The level of detail can be represented by D. The larger the D value, the higher the level of detail of the address. Let D(a) be the level of detail of address a, then: Final address = argmax a∈{源1地址，源2地址，源3地址} D(a).

[0100] For example: Agent 1 provides "A Province B City C District D Center Building A";

[0101] Agents one to four provided “A Province B City C District D Center A Building 6th Floor”.

[0102] The final output is: "6th Floor, Building A, D Center, District C, City B, Province A".

[0103] Detailed information usually provides higher precision and reliability. Prioritizing this information can reduce the ambiguity of the final results and improve the accuracy of decision-making.

[0104] In addition, if the level of detail of Agent 1 and Agent 2 is the same, they can be further differentiated by introducing weight factors and reliability scores. Define the level of detail:

[0105] D (agent 1 address) = 1; D (agent 2 address) = 2; D (agent 3 address) = 2;

[0106] Since Agent 2 and Agent 3 have the same level of detail, other factors may need to be considered, such as the consistency of the address format, the reliability of the source, etc. Assuming that Agent 2's source is more reliable, Agent 2 can be given a higher score. Ultimately, the address with the highest comprehensive score is selected as the final address. S(Source 1 address) = w1·D(Source 1 address) + w2·R1; S(Source 2 address) = w1·D(Source 2 address) + w2·R2;

[0107] ; where w1 and w2 are weight factors, R1, R2, and R3 are respectively source 1S (source 3 address) = w1·D (source 3 address) + w2·R3

[0108] The reliability scores of source 1 (agent 1), source 2 (agent 2), and source 3 (agent 3). Final address = argmax a∈{源1地址，源2地址，源3地址，} S(a).

[0109] Assuming that Agent Two’s source is more reliable, Agent Two’s address will be preferred as the final output even if both are at the same level of detail.

[0110] The multi-agent system can be flexibly configured according to different application scenarios and is suitable for various types of enterprise information acquisition needs. Through multi-source verification and intelligent decision-making mechanisms, the high accuracy and credibility of the extracted information are ensured. The entire process is highly automated, reducing the need for manual intervention and significantly improving work efficiency. It provides users with a fast and convenient way to obtain the required information and improves the service experience. Cross-verifying information through multiple channels reduces the risk of information bias or errors caused by a single source, thereby better protecting users from potential misleading.

[0111] The fusion of multi-source data helps eliminate the deviation or error caused by a single data source and enhances the reliability and accuracy of system decisions. Through missing data processing and voting mechanisms, the system can make reasonable decisions when the data is incomplete. Through automated task allocation, data processing and fusion rules, the system can reduce manual intervention and thus improve work efficiency. The combination of multi-agents and multi-source data enables the system to better adapt to various complex and dynamic application scenarios.

[0112] In summary, the multi-source fusion agent system provides an efficient, accurate and reliable solution for collecting and integrating the real address information of enterprises from multiple online resources.

[0113] S130: Output the final address information.

[0114] In this embodiment, the final address information is output to the terminal for display.

[0115] In this embodiment, Agents 1 to 4 form a composite agent, which can simultaneously process and understand data from different modalities (text, image, sound, etc.) by integrating multiple data processing technologies (such as natural language processing, computer vision, sound recognition, etc.). Such cross-modal data fusion greatly improves its applicability in complex application scenarios. For example, in image recognition, it is possible to combine image content with text information for comprehensive analysis, realize multi-dimensional data retrieval and understanding, and improve the comprehensiveness and accuracy of information extraction.

[0116] Based on deep learning and algorithm optimization, the composite agent extracts and fuses key features from different data modalities to enhance the analysis depth and recognition accuracy of data. When dealing with complex scenarios, the agent can optimize the recognition effect through self-learning, especially in fields such as OCR (optical character recognition), which can achieve higher accuracy and robustness than traditional methods.

[0117] The composite agent greatly reduces human intervention in the process of data collection, cleaning and processing. Through the embedded natural language processing and automated decision-making mechanism, it can intelligently extract key information from unstructured data and realize automatic data supplementation and correction. For example, in address cleaning, it can automatically identify and correct wrong addresses to ensure the accuracy and consistency of data.

[0118] Composite agents have powerful data collection and processing capabilities, and can extract, clean, convert and integrate data from multiple heterogeneous systems and different sources to ensure data consistency, integrity and availability. This capability gives them great advantages in multi-system collaboration and cross-platform data circulation.

[0119] The composite intelligent agent can not only automatically process data, but also coordinate and optimize cross-departmental and cross-system workflows. Through automated task allocation and management, it ensures efficient collaboration among all links, breaks down information silos, and effectively improves overall work efficiency.

[0120] This method is equipped with a powerful rule engine that can make judgments and decisions based on complex business needs and support dynamic adjustment of business rules. This function can respond to changes in the business environment in real time, adapt to different application scenarios, and ensure that the system continues to operate efficiently.

[0121] The composite agent has the ability to automatically assign tasks to appropriate robots or humans according to task requirements, while tracking task progress and generating detailed reports. This not only improves the efficiency of task execution, but also achieves the rational use of resources.

[0122] This method has powerful interface integration capabilities and can seamlessly connect with a variety of enterprise application systems, ensuring smooth flow of data between systems and efficient connection of business processes.

[0123] The composite agent is based on a large-scale corpus and a powerful algorithm model, and has the ability to continuously optimize and iterate. Through continuous algorithm training and optimization, its data processing and analysis capabilities will continue to improve, further improving the accuracy and reliability of the algorithm, especially in areas such as address standardization and longitude and latitude queries.

[0124] The above-mentioned method for identifying the actual business address of an enterprise collects relevant data from multiple data sources to ensure comprehensive information coverage and avoid the problem of incomplete data coverage in traditional technologies. Then, by combining the language model module in the intelligent body, multiple candidate address information is extracted, solving the problem of messy information in traditional technologies. Using data fusion technology, multiple extraction results are integrated to improve the accuracy of address information, overcoming the defects of limited accuracy of existing methods. Through intelligent algorithm optimization, the system can update and correct address information in real time, solving the problem that traditional technologies cannot be updated in time. Finally, accurate, comprehensive and up-to-date final address information is output, which greatly improves the efficiency and reliability of business address identification.

[0125] Figure 3 FIG. 3 is a schematic block diagram of a device 300 for identifying the actual business address of an enterprise provided by an embodiment of the present invention. Figure 3 As shown, corresponding to the above enterprise actual business address identification method, the present invention also provides an enterprise actual business address identification device 300. The enterprise actual business address identification device 300 includes a unit for executing the above enterprise actual business address identification method, and the device can be configured in a server. Figure 3 The enterprise actual business address identification device 300 includes an extraction unit 301, a fusion unit 302 and an output unit 303.

[0126] The extraction unit 301 is used to collect relevant data from different data sources and extract address information to obtain multiple extraction results; the fusion unit 302 is used to fuse the multiple extraction results to obtain final address information; the output unit 303 is used to output the final address information.

[0127] In one embodiment, the extraction unit 301 includes:

[0128] The first extraction subunit is used to obtain the company name when collecting the address from the company's official website, search for the official link according to the company name, and locate and extract the text information containing the keyword from the website corresponding to the official link, and use the text information as the extraction result;

[0129] The second extraction subunit is used to obtain the company name when collecting the address from the recruitment platform, search for relevant recruitment information according to the company name, and locate and extract the text information containing the keywords from the relevant recruitment information, and use the text information as the extraction result;

[0130] The third extraction subunit is used to obtain the enterprise name or unified social credit code when collecting the address from the enterprise information query platform, search for relevant enterprise information according to the enterprise name or unified social credit code, and locate and extract text information containing keywords from the relevant enterprise information, and use the text information as the extraction result;

[0131] The fourth extraction subunit is used to search the official social media and public accounts of the enterprise when collecting addresses from social media and public accounts, and extract text information containing keywords, and use the text information as the extraction result.

[0132] In one embodiment, the fusion unit 302 is used to:

[0133] When all extraction results are inconsistent, a voting mechanism is used to determine the final address information; when there are multiple identical addresses in multiple extraction results, a voting mechanism is used to select the address with the most occurrences as the final address information.

[0134] When the high-level information in multiple extraction results is the same but the low-level information is different, the high-level information is retained and the low-level information is merged to form the final address information, wherein the high-level information includes the street name and the building name, and the low-level information includes the floor number.

[0135] When some of the multiple extraction results are ambiguous, the details supplement and consistency priority rules are used to determine the final address information.

[0136] When the quantities in multiple extraction results are the same, the final address information is determined according to the preset priority rules.

[0137] It should be noted that technicians in the relevant field can clearly understand that the specific implementation process of the above-mentioned enterprise actual business address identification device 300 and each unit can refer to the corresponding description in the aforementioned method embodiment, and for the convenience and brevity of description, it will not be repeated here.

[0138] The above-mentioned enterprise actual business address identification device 300 can be implemented in the form of a computer program. The computer program can be used in Figure 4 Runs on the computer device shown.

[0139] See also Figure 4 , Figure 4 5 is a schematic block diagram of a computer device provided in an embodiment of the present application. The computer device 500 may be a server, wherein the server may be an independent server or a server cluster composed of multiple servers.

[0140] See also Figure 4The computer device 500 includes a processor 502 , a memory and a network interface 505 connected via a system bus 501 , wherein the memory may include a non-volatile storage medium 503 and an internal memory 504 .

[0141] The non-volatile storage medium 503 can store an operating system 5031 and a computer program 5032. The computer program 5032 includes program instructions, and when the program instructions are executed, the processor 502 can execute a method for identifying the actual business address of an enterprise.

[0142] The processor 502 is used to provide computing and control capabilities to support the operation of the entire computer device 500 .

[0143] The internal memory 504 provides an environment for the operation of the computer program 5032 in the non-volatile storage medium 503. When the computer program 5032 is executed by the processor 502, the processor 502 can execute a method for identifying the actual business address of an enterprise.

[0144] The network interface 505 is used to communicate with other devices over the network. Figure 4 The structure shown in the figure is merely a block diagram of a portion of the structure related to the solution of the present application, and does not constitute a limitation on the computer device 500 to which the solution of the present application is applied. The specific computer device 500 may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0145] The processor 502 is used to run the computer program 5032 stored in the memory to implement the following steps:

[0146] Collect relevant data from different data sources and extract address information to obtain multiple extraction results; merge the multiple extraction results to obtain final address information; and output the final address information.

[0147] Among them, the data sources include corporate official websites, recruitment platforms, corporate information query platforms, social media and public accounts.

[0148] In one embodiment, when the processor 502 implements the step of collecting relevant data from different data sources and extracting address information to obtain multiple extraction results, the processor 502 specifically implements the following steps:

[0149] When collecting addresses from the official website of an enterprise, obtain the enterprise name, search for official links based on the enterprise name, locate and extract text information containing keywords from the website corresponding to the official link, and use the text information as the extraction result; when collecting addresses from a recruitment platform, obtain the enterprise name, search for relevant recruitment information based on the enterprise name, locate and extract text information containing keywords from the relevant recruitment information, and use the text information as the extraction result; when collecting addresses from an enterprise information query platform, obtain the enterprise name or a unified social credit code, search for relevant enterprise information based on the enterprise name or the unified social credit code, locate and extract text information containing keywords from the relevant enterprise information, and use the text information as the extraction result; when collecting addresses from social media and official accounts, search for the official social media and official accounts of the enterprise, and extract text information containing keywords, and use the text information as the extraction result.

[0150] In one embodiment, when the processor 502 implements the step of fusing the multiple extraction results to obtain the final address information, the processor 502 specifically implements the following steps:

[0151] When all extraction results are inconsistent, a voting mechanism is used to determine the final address information; when there are multiple identical addresses in multiple extraction results, a voting mechanism is used to select the address with the most occurrences as the final address information.

[0152] In one embodiment, when the processor 502 implements the step of fusing the multiple extraction results to obtain the final address information, the processor 502 specifically implements the following steps:

[0153] When the high-level information in multiple extraction results is the same but the low-level information is different, the high-level information is retained and the low-level information is merged to form the final address information, wherein the high-level information includes the street name and the building name, and the low-level information includes the floor number.

[0154] In one embodiment, when the processor 502 implements the step of collecting relevant data from different data sources and extracting address information to obtain multiple extraction results, the processor 502 specifically implements the following steps:

[0155] When some of the multiple extraction results are ambiguous, the details supplement and consistency priority rules are used to determine the final address information.

[0156] In one embodiment, when the processor 502 implements the step of fusing the multiple extraction results to obtain the final address information, the processor 502 specifically implements the following steps:

[0157] When the quantities in multiple extraction results are the same, the final address information is determined according to the preset priority rules.

[0158] It should be understood that in the embodiment of the present application, the processor 502 may be a central processing unit (CPU), and the processor 502 may also be other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. Among them, the general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc.

[0159] It can be understood by those skilled in the art that all or part of the processes in the method for implementing the above embodiment can be completed by instructing the relevant hardware through a computer program. The computer program includes program instructions, and the computer program can be stored in a storage medium, which is a computer-readable storage medium. The program instructions are executed by at least one processor in the computer system to implement the process steps of the embodiment of the above method.

[0160] Therefore, the present invention also provides a storage medium. The storage medium may be a computer-readable storage medium. The storage medium stores a computer program, wherein when the computer program is executed by a processor, the processor executes the following steps:

[0161] Collect relevant data from different data sources and extract address information to obtain multiple extraction results; merge the multiple extraction results to obtain final address information; and output the final address information.

[0162] Among them, the data sources include corporate official websites, recruitment platforms, corporate information query platforms, social media and public accounts.

[0163] In one embodiment, when the processor executes the computer program to implement the step of collecting relevant data from different data sources and extracting address information to obtain multiple extraction results, the processor specifically implements the following steps:

[0164] When collecting addresses from the official website of an enterprise, obtain the enterprise name, search for official links based on the enterprise name, locate and extract text information containing keywords from the website corresponding to the official link, and use the text information as the extraction result; when collecting addresses from a recruitment platform, obtain the enterprise name, search for relevant recruitment information based on the enterprise name, locate and extract text information containing keywords from the relevant recruitment information, and use the text information as the extraction result; when collecting addresses from an enterprise information query platform, obtain the enterprise name or a unified social credit code, search for relevant enterprise information based on the enterprise name or the unified social credit code, locate and extract text information containing keywords from the relevant enterprise information, and use the text information as the extraction result; when collecting addresses from social media and official accounts, search for the official social media and official accounts of the enterprise, and extract text information containing keywords, and use the text information as the extraction result.

[0165] In one embodiment, when the processor executes the computer program to implement the step of fusing the multiple extraction results to obtain the final address information, the processor specifically implements the following steps:

[0166] When all extraction results are inconsistent, a voting mechanism is used to determine the final address information; when there are multiple identical addresses in multiple extraction results, a voting mechanism is used to select the address with the most occurrences as the final address information.

[0167] In one embodiment, when the processor executes the computer program to implement the step of fusing the multiple extraction results to obtain the final address information, the processor specifically implements the following steps:

[0168] When the high-level information in multiple extraction results is the same but the low-level information is different, the high-level information is retained and the low-level information is merged to form the final address information, wherein the high-level information includes the street name and the building name, and the low-level information includes the floor number.

[0169] In one embodiment, when the processor executes the computer program to implement the step of collecting relevant data from different data sources and extracting address information to obtain multiple extraction results, the processor specifically implements the following steps:

[0170] When some of the multiple extraction results are ambiguous, the details supplement and consistency priority rules are used to determine the final address information.

[0171] In one embodiment, when the processor executes the computer program to implement the step of fusing the multiple extraction results to obtain the final address information, the processor specifically implements the following steps:

[0172] When the quantities in multiple extraction results are the same, the final address information is determined according to the preset priority rules.

[0173] The storage medium may be a USB flash drive, a mobile hard disk, a read-only memory (ROM), a magnetic disk, or an optical disk, etc., which are computer-readable storage media that can store program codes.

[0174] Those of ordinary skill in the art will appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein can be implemented in electronic hardware, computer software, or a combination of the two. In order to clearly illustrate the interchangeability of hardware and software, the composition and steps of each example have been generally described in terms of function in the above description. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional and technical personnel can use different methods to implement the described functions for each specific application, but such implementation should not be considered to be beyond the scope of the present invention.

[0175] In the several embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of each unit is only a logical function division, and there may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed.

[0176] The steps in the method of the embodiment of the present invention can be adjusted in order, combined and deleted according to actual needs. The units in the device of the embodiment of the present invention can be combined, divided and deleted according to actual needs. In addition, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit.

[0177] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a storage medium. Based on this understanding, the technical solution of the present invention is essentially or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product, and the computer software product is stored in a storage medium, including a number of instructions for a computer device (which can be a personal computer, terminal, or network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention.

[0178] The above is only a specific embodiment 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 identifying the actual business address of an enterprise, characterized in that: include: Collect relevant data from different data sources and extract address information to obtain multiple extraction results; Merging the multiple extraction results to obtain final address information; The final address information is output.

2. The method for identifying the actual business address of an enterprise according to claim 1, characterized in that: The data sources include corporate official websites, recruitment platforms, corporate information query platforms, social media and public accounts.

3. The method for identifying the actual business address of an enterprise according to claim 2, characterized in that: The relevant data is collected from different data sources, and address information is extracted to obtain multiple extraction results, including: When collecting addresses from the official website of an enterprise, obtain the enterprise name, search for official links based on the enterprise name, and locate and extract text information containing keywords from the website corresponding to the official link, and use the text information as the extraction result; When collecting addresses from the recruitment platform, obtain the company name, search for relevant recruitment information based on the company name, and locate and extract text information containing keywords from the relevant recruitment information, and use the text information as the extraction result; When collecting addresses from the enterprise information query platform, obtaining the enterprise name or unified social credit code, searching for relevant enterprise information based on the enterprise name or unified social credit code, and locating and extracting text information containing keywords from the relevant enterprise information, and using the text information as the extraction result; When collecting addresses from social media and official accounts, search for the official social media and official accounts of the enterprise, and extract text information containing keywords, and use the text information as the extraction result.

4. The method for identifying the actual business address of an enterprise according to claim 3, characterized in that: The fusing of the multiple extraction results to obtain final address information includes: When all extraction results are inconsistent, a voting mechanism is used to determine the final address information; when there are multiple identical addresses in multiple extraction results, a voting mechanism is used to select the address with the most occurrences as the final address information.

5. The method for identifying the actual business address of an enterprise according to claim 3, characterized in that: The fusing of the multiple extraction results to obtain final address information includes: When the high-level information in multiple extraction results is the same but the low-level information is different, the high-level information is retained and the low-level information is merged to form the final address information, wherein the high-level information includes the street name and the building name, and the low-level information includes the floor number.

6. The method for identifying the actual business address of an enterprise according to claim 3, characterized in that: The relevant data is collected from different data sources, and address information is extracted to obtain multiple extraction results, including: When some of the multiple extraction results are ambiguous, the details supplement and consistency priority rules are used to determine the final address information.

7. The method for identifying the actual business address of an enterprise according to claim 3, characterized in that: The fusing of the multiple extraction results to obtain final address information further includes: When the quantities in multiple extraction results are the same, the final address information is determined according to the preset priority rules.

8. The device for identifying the actual business address of an enterprise is characterized by: include: An extraction unit, used to collect relevant data from different data sources and extract address information to obtain multiple extraction results; A fusion unit, used for fusing the multiple extraction results to obtain final address information; An output unit is used to output the final address information.

9. A computer device, characterized in that: The computer device comprises a memory and a processor, wherein a computer program is stored in the memory, and when the processor executes the computer program, the method according to any one of claims 1 to 7 is implemented.

10. A storage medium, characterized in that: The storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 7 is implemented.

Images