Resume document matching method, device, computing device and storage medium

Abstract

The present application provides a resume document matching method, device, computing device and storage medium, wherein the resume person-job matching method includes: obtaining a target resume document to be matched and a preset job description; extracting text content in the target resume document; determining the semantic similarity between the target resume document and the job description based on the text content and the job description; and determining the matching result between the target resume document and the job description based on the semantic similarity. Through the above method, the semantic similarity between the target resume document and the job description is determined based on the text content and the job description extracted from the target resume document, the semantic information between the target resume document and the job description is mined, and the resume document is matched based on the semantic similarity between the target resume document and the job description, thereby improving the accuracy of the matching result.

Description

Technical Field

[0001] The present application relates to the field of computer technology, and more particularly to a resume document matching method, a resume document matching device, a computing device, and a computer-readable storage medium. Background Art

[0002] With the development of Internet technology, automated resume document matching has been widely used in recruitment, greatly facilitating the recruitment efficiency of job seekers and employers. Resume document matching refers to matching the resume document of job seekers with the employment requirements published by employers. It can recommend the most suitable positions based on the resume document of job seekers, and can also recommend the most suitable job seekers based on employment requirements.

[0003] Currently, resume document matching usually involves extracting keywords from the resume document of a job seeker and the hiring needs of an employer respectively, and then matching the extracted keywords to determine whether the resume document meets the hiring needs of the employer based on the matching results.

[0004] However, matching based on keywords can only match keywords with exactly the same characters, which has poor applicability. In practical applications, the accuracy of resume document matching results is low. Summary of the invention

[0005] In view of this, the embodiment of the present application provides a resume document matching method to solve the technical defects existing in the prior art. The embodiment of the present application also provides a resume document matching device, a computing device, and a computer-readable storage medium.

[0006] According to a first aspect of an embodiment of the present application, a resume document matching method is provided, comprising:

[0007] Obtain the target resume document to be matched and the preset job description;

[0008] Extract text content from the target resume document;

[0009] Determine the semantic similarity between the target resume document and the job description based on the text content and the job description;

[0010] Based on semantic similarity, the matching results between the target resume document and the job description are determined.

[0011] According to a second aspect of an embodiment of the present application, a resume document matching device is provided, comprising:

[0012] An acquisition module is configured to acquire a target resume document to be matched and a preset job description;

[0013] An extraction module, configured to extract text content from a target resume document;

[0014] a determination module configured to determine semantic similarity between a target resume document and the job description based on text content and the job description;

[0015] The matching module is configured to determine the matching result between the target resume document and the job description based on semantic similarity.

[0016] According to a third aspect of an embodiment of the present application, a computing device is provided, including:

[0017] Memory and processor;

[0018] The memory is used to store computer-executable instructions, and the processor implements the steps of the resume document matching method when executing the computer-executable instructions.

[0019] According to a fourth aspect of an embodiment of the present application, a computer-readable storage medium is provided, which stores computer-executable instructions, and when the instructions are executed by a processor, the steps of the resume document matching method are implemented.

[0020] The resume document matching method provided in the present application obtains the target resume document to be matched and the preset job description, then extracts the text content in the target resume document, and determines the semantic similarity between the target resume document and the job description based on the extracted text content and the job description. Based on the semantic similarity, the matching result of the target resume document and the job description can be determined. Through the above method, the semantic similarity between the target resume document and the job description is determined based on the text content and the job description extracted from the target resume document, the semantic information between the target resume document and the job description is mined, and the resume document is matched based on the semantic similarity between the target resume document and the job description, thereby improving the accuracy of the matching result. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 A schematic diagram of the structure of a resume document matching system provided according to an embodiment of the present application is shown;

[0022] Figure 2 A flow chart of a resume document matching method provided according to an embodiment of the present application is shown;

[0023] Figure 3 A flow chart of a method for extracting text content from a target resume document provided according to an embodiment of the present application is shown;

[0024] Figure 4 A flow chart of a method for extracting text content from a job description provided according to an embodiment of the present application is shown;

[0025] Figure 5 A flow chart of a method for determining semantic similarity between a target resume document and a job description provided according to an embodiment of the present application is shown;

[0026] Figure 6 A flow chart of a method for determining semantic similarity between a target resume document and a corresponding part in a job description provided according to an embodiment of the present application is shown;

[0027] Figure 7 A flow chart of a method for matching a target resume document with a job description provided according to an embodiment of the present application is shown;

[0028] Figure 8 A flow chart of a method for determining a target position provided according to an embodiment of the present application is shown;

[0029] Fig. 9 A flow chart of a method for recommending content to be added to a resume according to an embodiment of the present application is shown;

[0030] Fig.10 A flow chart of a method for correcting a target resume document provided according to an embodiment of the present application is shown;

[0031] Fig.11 A flow chart of a method for detecting an incorrect sentence according to an embodiment of the present application is shown;

[0032] Fig.12 A flowchart of another resume document matching method provided according to an embodiment of the present application is shown;

[0033] Fig.13 A schematic diagram of the structure of a resume document matching device provided in one embodiment of the present application is shown;

[0034] Fig.14 A structural block diagram of a computing device provided according to an embodiment of the present application is shown. DETAILED DESCRIPTION

[0035] Many specific details are described in the following description to facilitate a full understanding of the present application. However, the present application can be implemented in many other ways than those described herein, and those skilled in the art can make similar generalizations without violating the connotation of the present application, so the present application is not limited by the specific implementation disclosed below.

[0036] The terms used in one or more embodiments of the present application are only for the purpose of describing specific embodiments, and are not intended to limit one or more embodiments of the present application. The singular forms of "a", "said" and "the" used in one or more embodiments of the present application and the appended claims are also intended to include plural forms, unless the context clearly indicates other meanings. It should also be understood that the term "and / or" used in one or more embodiments of the present application refers to and includes any or all possible combinations of one or more associated listed items.

[0037] It should be understood that, although the terms first, second, etc. may be used to describe various information in one or more embodiments of the present application, such information should not be limited to these terms. These terms are only used to distinguish the same type of information from each other. For example, without departing from the scope of one or more embodiments of the present application, the first may also be referred to as the second, and similarly, the second may also be referred to as the first.

[0038] First, the terms involved in one or more embodiments of the present invention are explained.

[0039] Bidirectional Encoder Representation from Transformers (BERT) model: It is a pre-trained language model. Through a large amount of data training, the BERT model can extract relational features at multiple different levels, obtain word meanings based on the sentence context, avoid ambiguity, and thus more comprehensively reflect the semantics of sentences.

[0040] Job description (JD): The employer's specific description of the position, job duties, and responsibilities of the job.

[0041] String method: is a method to convert natural language into string data, which is more conducive to machine recognition and improves the efficiency of machine text recognition.

[0042] In the present application, a resume document matching method is provided. The present application also relates to a resume document matching device, a computing device, and a computer-readable storage medium, which are described in detail one by one in the following embodiments.

[0043] Figure 1 A schematic diagram of the structure of a resume document matching system provided according to an embodiment of the present application is shown.

[0044] The resume document matching system of the present application includes a training end and an application end. The training end trains a semantic similarity model and then sends the trained semantic similarity model to the application end. The application end is used to provide samples to the training end and receive the model trained by the training end, and use the model to match the target resume document to be matched with the job description to obtain a matching result. The application end and the training end can be different applications in a terminal device or applications in different terminal devices.

[0045] In this embodiment, the training end can train the semantic similarity model by the following method: select a large amount of resume document data (such as basic information, job-seeking intention, work experience, educational experience, scientific research direction, academic achievements, self-evaluation, etc.) and job description data (such as recruitment objects, school types, graduation time, professional ability, project experience, etc.) as the training sample set of the BERT model, randomly mask a part of the resume document data and job description data, and adjust the parameters of the model to enable the BERT model to predict the masked resume document data and job description data, and then finally reach the preset threshold after multiple iterations of training, and use the BERT model that reaches the preset threshold as the trained BERT model.

[0046] The application side can match the target resume document by the following method: obtain the target resume document to be matched and the preset job description, then extract the text content in the target resume document, and determine the semantic similarity between the target resume document and the job description based on the text content and job description using the semantic similarity model trained on the training side, and finally determine the matching result between the target resume document and the job description based on the semantic similarity.

[0047] In addition, the application end can also match the target resume document by the following method: obtain the target resume document to be matched and the preset job description, then extract the text content in the target resume document, and use an unsupervised algorithm (such as term frequency-inverse text frequency (TF-IDF, Term Frequency-Inverse Document Frequency), Euclidean distance method, etc.) to determine the semantic similarity between the target resume document and the job description, and finally determine the matching result between the target resume document and the job description based on the semantic similarity. This application does not limit the method for matching the target resume document.

[0048] Through the above method, the semantic similarity between the target resume document and the job description is determined based on the text content extracted from the target resume document and the preset job description, the semantic information between the target resume document and the job description is mined, and resume document matching is performed based on the semantic similarity between the target resume document and the job description, thereby improving the accuracy of the matching results.

[0049] Figure 2 A flow chart of a resume document matching method provided according to an embodiment of the present application is shown, which specifically includes the following steps:

[0050] S202: Obtain a target resume document to be matched and a preset job description.

[0051] The execution subject of the resume document matching method can be a computing device with a resume document matching function, such as a server, terminal, etc. with a resume document matching function.

[0052] The target resume document refers to the resume document edited by the applicant, which may include basic information, job search intention, work experience, educational experience, research direction, academic achievements, self-evaluation, etc. The job description is the employer's specific description of the position, job duties, job responsibilities, etc. of the recruitment position, which may include the recruitment object, school type, graduation time, professional ability, project experience, etc.

[0053] The target resume document and job description can be in the format of text, image, etc. There are many ways to obtain the target resume document and job description. For example, the user can send an instruction to obtain the target resume document and job description. Correspondingly, after receiving the instruction, the target resume document and job description are obtained. It can also be that the specified target resume document and job description to be matched are automatically obtained at preset time intervals. This application does not impose any restrictions on the method of obtaining the target resume document and job description to be matched.

[0054] S204: extracting text content from the target resume document.

[0055] After obtaining the target resume document to be matched, it is necessary to extract the text content of string type from the target resume document. There are many ways to extract the text content of string type from the target document. For example, the String method can be used to extract the text content in the target resume document; the StringBuffer method or the StringBuilder method can also be used to extract the text content in the target resume document. In a preferred implementation, the String method is used to extract the text content in the target resume document. The purpose of using the String method to extract text content is to convert text in natural language form into text in string form, which is more conducive to machine recognition and improves the efficiency of resume document matching.

[0056] In a possible implementation method of the embodiment of the present application, it is identified that the format of the target resume document is a text format, and the text content in the target resume document is extracted, and then the String method is used to convert the text in natural language form into text content in string form, wherein the text content in the target resume document can be extracted by extracting all the text content in the target resume document, or by extracting part of the text content according to preset extraction rules, for example, only extracting the text content about education experience, work experience, project experience, etc. in the target resume document.

[0057] In another possible implementation of the embodiment of the present application, if the target resume document is identified as a picture format, optical character recognition (OCR) technology is used to extract text from the target resume document. For example, for a target resume document in PDF format, OCR technology is used to identify the text area, and then the text area is segmented into different characters by rectangles, and then the characters are classified to identify the text content. The identified text is converted into text content of string type using the String method.

[0058] S206: Determine the semantic similarity between the target resume document and the job description based on the text content and the job description.

[0059] In one implementation of an embodiment of the present application, the semantic similarity between the target resume document and the job description can be determined by a pre-trained semantic similarity model. The semantic similarity model can be a BERT model. The BERT model is a pre-trained language model. Through a large amount of data training, the model can extract relational features at multiple different levels, obtain word meanings based on the sentence context, avoid the emergence of ambiguity, and thus more comprehensively reflect the semantics of the sentence. It should be noted that the semantic similarity model can not only be a BERT model, any neural network model whose input is a pair of sentences and whose output is a similarity score between sentences can be used as the semantic similarity model of the present application. The present application does not impose any restrictions on the type of semantic similarity model.

[0060] The BERT model is pre-trained. For example, a large amount of resume document data (such as basic information, job-seeking intentions, work experience, educational experience, scientific research directions, academic achievements, self-evaluation, etc.) and job description data (such as recruitment targets, school types, graduation time, professional capabilities, project experience, etc.) are selected as the training sample set of the BERT model. A part of the resume document data and job description data are randomly masked. By adjusting the parameters of the model, the BERT model is used to predict the masked resume document data and job description data. After multiple iterations of training, the preset threshold is finally reached. The BERT model that reaches the preset threshold is used as the trained BERT model.

[0061] In a possible implementation of an embodiment of the present application, after extracting the text content in the target resume document, the text content and the preset job description are input into the BERT model to determine the semantic similarity between the target resume document and the job description.

[0062] Of course, using the BERT model to determine the semantic similarity between the target resume document and the job description is only one implementation method. The semantic similarity can also be determined through feature matching, character comparison, etc. The BERT model, as a deep learning model, has high accuracy and calculation speed, so it can be adopted as a preferred embodiment in the embodiments of the present application.

[0063] S208. Determine the matching result between the target resume document and the job description based on the semantic similarity.

[0064] The matching result between the target resume document and the job description determined based on the semantic similarity represents the degree of matching between the target resume document and the job description. The higher the semantic similarity, the higher the matching degree between the target resume document and the job description. If the semantic similarity between the target resume document and the job description reaches the preset threshold, it means that the target resume document and the job description are successfully matched; if the semantic similarity between the target resume document and the job description does not reach the preset threshold, it means that the target resume document and the job description fail to match.

[0065] In a possible implementation method of an embodiment of the present application, the semantic similarity threshold between the target resume document and the job description is pre-set to 90, the text content of the target resume document and the job description are input into the trained BERT model, and the similarity between the text content of the target resume document and the job description is obtained to be 95, then it is determined that the target resume document and the job description are successfully matched.

[0066] In the above implementation, the target resume document to be matched and the preset job description are obtained, and then the text content in the target resume document is extracted. According to the extracted text content and job description, the semantic similarity between the target resume document and the job description is determined. Based on the semantic similarity, the matching result of the target resume document and the job description can be determined. Through the above method, according to the text content and job description extracted from the target resume document, the semantic similarity between the target resume document and the job description is determined, the semantic information between the target resume document and the job description is mined, and the resume document is matched according to the semantic similarity between the target resume document and the job description, thereby improving the accuracy of the matching result.

[0067] exist Figure 2 In the illustrated embodiment, when extracting the text content in the target resume document, the target resume document may be subjected to standardization processing. Accordingly, the steps of extracting the text content in the target resume document are as follows: Figure 3 As shown, Figure 3 A flowchart of a method for extracting text content from a target resume document provided according to an embodiment of the present application is shown, which specifically includes the following steps:

[0068] S302: Based on the specified resume format, standardize the target resume document.

[0069] S304, extracting the text content of each part of the target resume document after standardization processing.

[0070] Specifying a resume format refers to dividing different parts according to different contents in a resume document. For example, a resume format can be divided into a basic information part, an education experience part, a work experience part, etc. Standardizing a target resume document refers to classifying the contents of the target resume document according to the specified resume format.

[0071] Due to the personalized factors of the target resume document, the resume document layout and content description of each applicant are different. Usually, the target resume document that has not been standardized contains a lot of useless information. If the applicant's resume document is directly matched with the job description for semantic similarity, the matching result will be inaccurate, reducing the accuracy of resume document matching. Therefore, it is necessary to standardize the target resume document according to the specified resume format, which can not only filter out useless information, but also provide a unified resume document format for different target resume documents, thereby improving the accuracy of target resume document matching results.

[0072] There are many ways to standardize the target resume document. You can delete useless information such as spaces and unrecognizable characters; you can also pre-build keywords corresponding to each part of the resume format in the database and match the text content in the target resume document with the keywords; you can also match the target resume document based on a pre-established keyword comparison table, etc.

[0073] In one implementation of the present application, the target resume document is standardized by pre-building keywords corresponding to each part of the resume document in the database and matching the text content in the target resume document with the keywords. Table 1 schematically lists the keywords for standardizing the target resume document.

[0074] Table 1 Keyword list for standardized processing of resume documents

[0075]

[0076]

[0077] Taking Table 1 as an example, based on the four parts of the resume document, four databases are pre-constructed, namely the basic information database, the education experience database, the work experience database and the skill specialty database. Among them, the basic information database contains keywords corresponding to the basic information part, such as "basic information, basic information, personal information, personal profile, personal introduction, personal summary, basic information, age, gender, identity information", etc.; the education experience database contains keywords corresponding to the education experience part, such as "graduation school, school experience, on-campus experience, club activities, student work, educational practice, scientific research papers, scientific research directions, academic achievements, awards, professional ability, major courses, on-campus positions, practical research", etc.; the work experience database contains keywords corresponding to the work experience part, such as "work background, work experience, professional experience, project introduction, project experience", etc.; the skill specialty database contains keywords corresponding to the skill specialty part, such as "C language, Java, C++", etc.

[0078] In a possible implementation of the embodiment of the present application, when the target resume document to be matched is obtained, the text content of the target resume document is read to include "basic information, professional ability, major courses, age, gender, graduation school, professional experience, C language", and the text content of the target resume document is standardized according to the keywords corresponding to the resume format specified in Table 1, and the result of the standardized processing is: "Basic information part: basic information, age, gender; education experience part: graduation school, major courses, professional ability; work experience part: professional experience; skills and expertise part: C language". Then, the text content of the basic information part, education experience part, and work experience part of the standardized target resume document is extracted in turn.

[0079] In the above implementation, the target resume document is standardized based on the specified resume format, and then the text content of each part of the standardized target resume document is extracted, so that a unified resume document format can be provided for different target resume documents, thereby improving the accuracy of the target resume document matching results.

[0080] based on Figure 2 , Figure 3 In the embodiment shown, in addition to extracting text content from the target resume document, in order to facilitate the comparison between the target resume document and the job description, it is also necessary to extract text content from the job description. Accordingly, the steps of extracting text content from the job description are as follows: Figure 4 As shown, Figure 4 A flowchart of a method for extracting text content from a job description provided according to an embodiment of the present application is shown, which specifically includes the following steps:

[0081] S402: Standardize the job description based on the specified resume format.

[0082] S404, extracting the text content of each part of the job description after standardization.

[0083] same Figure 3 In the embodiment shown, the specified resume format refers to different parts divided according to different contents in the job description. For example, the resume format can be divided into a basic information part, an education experience part, a work experience part, etc. The specific standardized processing method can be the same as Figure 3 The method in the illustrated embodiment is the same and will not be described again here.

[0084] In a first possible implementation method of an embodiment of the present application, after obtaining the target resume document to be matched and the preset job description, the text content in the preset job description is matched with the keywords corresponding to each part of the pre-constructed resume format, that is, only the preset job description is standardized.

[0085] In a second possible implementation method of the embodiment of the present application, after obtaining the target resume document and the preset job description to be matched, the text content in the target resume document and the preset job description is matched with the keywords corresponding to the various parts of the pre-constructed resume format, that is, the target resume document and the preset job description are standardized respectively.

[0086] In the above implementation, the preset job description is standardized based on the specified resume format, and then the text content of each part of the preset job description after standardization is extracted, so that a unified resume document format can be provided for different job descriptions, thereby improving the accuracy of the matching results between the target resume document and the preset job description.

[0087] exist Figure 2 In the embodiment shown, when determining the semantic similarity between the target resume document and the job description, the specific method used can be as follows: Figure 5 As shown, Figure 5 A flow chart of a method for determining the semantic similarity between a target resume document and a job description according to an embodiment of the present application is shown, which specifically includes the following steps:

[0088] S502: Determine the semantic similarity between each part in the target resume document and the corresponding part in the job description based on the text content of each part in the target resume document and the text content of the corresponding part in the job description.

[0089] S504: Weight each semantic similarity to obtain the semantic similarity between the target resume document and the job description.

[0090] In this embodiment, after extracting the text contents of each part of the target resume document and the job description in turn, the semantic similarity of the text contents of each part of the target resume document and the text of the corresponding part of the job description is calculated respectively to determine the similarity of the text contents of each part of the target resume document and the text of the corresponding part of the job description, and then each semantic similarity is weighted to obtain the semantic similarity between the target resume document and the job description.

[0091] In a first possible implementation of the embodiment of the present application, after obtaining the target resume document to be matched, the text content in the target resume document and the preset job description is matched with the keywords by using the keywords corresponding to each part of the pre-constructed resume format, that is, the target resume document and the preset job description are standardized respectively. The text content of the target resume document is standardized according to the keywords corresponding to the resume format specified in Table 1, and the result of the standardized processing is: "Basic information part: basic information, age, gender; education experience part: graduation school, major courses, professional ability; work experience part: professional experience; skills and expertise part: C language".

[0092] The text content of the job description was standardized according to the keywords corresponding to the resume format specified in Table 1, and the standardized results were: "Basic information part: basic information, personal profile; Education experience part: graduation school, major courses, academic achievements; Work experience part: professional experience; Skills and expertise part: C language, Java, C++.

[0093] Then, the text contents of the basic information part, education experience part, work experience part, and skill specialty part of the standardized target resume document and job description are extracted respectively, and the similarity between each part is determined. The similarity between each part is weighted to obtain the semantic similarity between the target resume document and the job description. If the semantic similarity between the target resume document and the job description reaches a preset threshold, it is determined that the target resume document and the job description are successfully matched; if the semantic similarity between the target resume document and the job description does not reach the preset threshold, it is determined that the target resume document and the job description fail to match.

[0094] In a second possible implementation method of the embodiment of the present application, after obtaining the target resume document to be matched and the preset job description, the text content in the target resume document is matched with the keywords corresponding to each part of the resume format pre-constructed in the server database, that is, only the target resume document is standardized.

[0095] Then, the text contents of the basic information part, education experience part, work experience part, and skills and expertise part of the standardized target resume document and the unstandardized job description are extracted respectively, and the similarity between each part is determined. The similarity between each part is weighted to obtain the semantic similarity between the target resume document and the job description. If the semantic similarity between the target resume document and the job description reaches a preset threshold, it is determined that the target resume document and the job description are successfully matched; if the semantic similarity between the target resume document and the job description does not reach the preset threshold, it is determined that the target resume document and the job description fail to match.

[0096] In a third possible implementation method of the embodiment of the present application, after obtaining the target resume document to be matched and the preset job description, the text content of the preset job description is matched with the keywords corresponding to each part of the resume format pre-constructed in the server database, that is, only the preset job description is standardized.

[0097] Then, the text contents of the basic information part, education experience part, work experience part, and skills and expertise part of the target resume document that has not been standardized and the job description that has been standardized are extracted respectively, and the similarity between each part is determined. The similarity between each part is weighted to obtain the semantic similarity between the target resume document and the job description. If the semantic similarity between the target resume document and the job description reaches a preset threshold, it is determined that the target resume document and the job description are successfully matched; if the semantic similarity between the target resume document and the job description does not reach the preset threshold, it is determined that the target resume document and the job description fail to match.

[0098] In a fourth possible implementation method of the embodiment of the present application, after obtaining the target resume document to be matched and the preset job description, the text content of the basic information part, education experience part, work experience part, and skill specialty part in the target resume document and the job description is directly extracted, and the similarity between the parts is determined, and the similarity between the parts is weighted to obtain the semantic similarity between the target resume document and the job description. If the semantic similarity between the target resume document and the job description reaches a preset threshold, it is determined that the target resume document and the job description are matched successfully; if the semantic similarity between the target resume document and the job description does not reach the preset threshold, it is determined that the target resume document and the job description fail to match.

[0099] In the above implementation, based on the text content of each part in the target resume document and the text content of the corresponding part in the job description, the semantic similarity between each part in the target resume document and the corresponding part in the job description can be determined, and then each semantic similarity is weighted to obtain the semantic similarity between the target resume document and the job description. Based on the semantic similarity between the target resume document and the job description, the accuracy of the matching result of the target resume document can be improved.

[0100] exist Figure 5 In the embodiment shown, the target resume document may specifically include a work experience section; the job description may specifically include a job requirements section. When determining the semantic similarity between the target resume document and the corresponding section in the job description, the specific method used may be as follows: Figure 6 As shown, Figure 6A flow chart of a method for determining the semantic similarity between a target resume document and a corresponding part in a job description according to an embodiment of the present application is shown, which specifically includes the following steps:

[0101] S602: Match the text content of the work experience part with the text content of the job requirements part to obtain the degree of overlap of the work experience.

[0102] S604. If the work experience overlap is greater than a preset threshold, the text content of the work experience part and the text content of the job requirements part are input into a pre-trained semantic similarity model to obtain the experience similarity between the work experience part and the job requirements part.

[0103] The work experience overlap refers to the percentage of overlap between the work experience part and the job requirements part. The purpose of the work experience overlap is to determine whether it is necessary to calculate the semantic similarity between the work experience part and the job requirements part. If the work experience overlap is greater than the preset threshold, the text content of the work experience part and the text content of the job requirements part are input into the semantic similarity model; if the work experience overlap is less than the preset threshold, it is directly determined that the work experience part in the target resume document and the job requirements part in the job description fail to match, thereby reducing the calculation amount of the semantic similarity model and improving the matching efficiency of the target resume document.

[0104] In this embodiment, based on the specified resume format, the text content of the work experience section in the target resume document is matched with the text content of the job requirements section in the job description by keywords. If the overlap percentage of the keyword matching between the work experience section and the job requirements section is greater than a preset threshold, the keywords of the work experience section and the job requirements section are input into the semantic similarity model together to obtain the experience similarity between the work experience section and the job requirements section.

[0105] In a possible implementation of the embodiment of the present application, the semantic similarity model can be a BERT model. The BERT model is a pre-trained language model. Through a large amount of data training, the model can extract relational features at multiple different levels, obtain word meanings according to the context of the sentence, avoid the appearance of ambiguity, and thus more comprehensively reflect the semantics of the sentence. Taking the preset threshold of work experience overlap as 30% as an example, the keywords of the text content of the work experience part in the target resume document after standardization are extracted as "artificial intelligence, programming design, C language", and the keywords of the text content of the job requirements part in the job description after standardization are extracted as "artificial intelligence, Java, C language, C++", and the keywords in the text content of the work experience part are matched with the keywords in the text of the job requirements part, and the work experience overlap is obtained. The overlap is 50%, which is greater than the preset work experience overlap threshold. The text content of the work experience part and the text content of the job requirements part are input into the pre-trained BERT model, and then the experience similarity score of the work experience part and the job requirements part is obtained.

[0106] Through the above method, the text content of the work experience part is first matched with the text content of the job requirements part to obtain the work experience overlap, so as to determine whether it is necessary to perform semantic similarity calculation on the work experience part and the job requirements part, thereby reducing the calculation amount of the semantic similarity model and improving the matching efficiency of the target resume document; the work experience overlap is combined with the semantic similarity model to improve the accuracy of the matching result of the target resume document.

[0107] Of course, in some implementations, the text content of a certain part of the target resume document can also be matched with the text content of the corresponding part of the job description (the corresponding relationship is generally pre-set) to obtain the text content overlap. Only when the text content overlap is greater than a certain threshold will the step of matching the target resume document with the job description be continued, which can also improve the accuracy of the target resume document matching results.

[0108] based on Figure 6 In the embodiment shown, the target resume document may further include an education experience section and a job requirement section; the job description may further include a job position section. In addition to calculating the experience similarity, the job position similarity may also be calculated. Accordingly, the target resume document may be matched with the job description in the following manner: Figure 7 As shown, Figure 7 A flow chart of a method for matching a target resume document with a job description according to an embodiment of the present application is shown, which specifically includes the following steps:

[0109] S702: Input the text content of the job requirement part and the text content of the job position part into a semantic similarity model to obtain the job similarity between the job requirement part and the job position part.

[0110] S704, extract key information and additional information from the education experience part and the work experience part respectively.

[0111] S706: Allocate a first score based on the matching result between the key information and the preset key information condition.

[0112] S708: Allocate a second score based on the matching result between the additional information and the preset additional information condition.

[0113] S710: Weight the experience similarity, position similarity, the first score, and the second score to obtain a matching degree between the target resume document and the position description.

[0114] S712. Determine the matching result between the target resume document and the job description according to the matching degree.

[0115] The target resume document also includes an education experience section and a job requirement section, where the education experience section refers to the applicant's educational background; the job requirement section refers to the position the applicant requires. The job description includes a position section, where the position section refers to the position the recruiter is recruiting for. Key information refers to the basic requirements of the recruiter for the applicant. Additional information refers to the bonus points preset by the recruiter for the applicant, such as passing the CET-6 exam, honorary certificates, familiarity with the Python language, etc. The more additional information the applicant meets, the higher the success rate of his resume document matching.

[0116] After obtaining the position similarity by comparing the similarity between the position requirement part and the position part in the position description in the target resume document, it is necessary to extract key information and additional information from the education experience part and the work experience part respectively. There are many ways to extract key information and additional information from the education experience part and the work experience part respectively. You can extract key information from the education experience part and extract additional information from the work experience part; you can also extract additional information from the education experience part and extract key information from the work experience part; you can also extract key information and additional information from the education experience part, and extract key information and additional information from the work experience part. Of course, you can also integrate the education experience part and the work experience part first, and then extract key information and experience information from the integrated content, that is, the key information contains the corresponding content of the education experience part and the work experience part, and the additional information contains the corresponding content of the education experience part and the work experience part. The integrated content is compared with the key information database and the additional information database respectively, and the key information and additional information can be extracted.

[0117] In an embodiment of the present application, two keyword databases are pre-constructed, namely a key information database and an additional information database, wherein the key information database contains keywords corresponding to the key information, such as "982 / 211 colleges, undergraduate, computer, software, programming"; the additional information database contains keywords corresponding to the additional information, such as "C language, Python language, CET-6, honor certificate, project experience" and the like.

[0118] After extracting key information and additional information from the education experience section and the work experience section, respectively, the key information of the target resume document is matched with the preset key information conditions to obtain a first score, which represents the degree of matching between the key information and the preset key information conditions. The higher the matching degree (for example, judging the number of identical words; or judging the overlap in the order of matching education experience first and then matching work experience), the higher the first score is assigned. The additional information extracted from the education experience section and the work experience section of the target resume document is matched with the preset additional information conditions to obtain a second score, which represents the degree of matching between the additional information and the preset additional information conditions. The higher the matching degree (for example, judging the number of identical words; or judging the overlap in the order of the additional skill requirements), the higher the second score is assigned.

[0119] In a specific implementation, key information can be extracted from the education experience part and the work experience part respectively. Therefore, a first score can be calculated for the education experience part, and another first score can be calculated for the work experience part. Then, in this case, the final first score can be calculated based on the two first scores for the education experience part and the work experience part. The specific calculation method can be to find the average of the two, or to assign different weights according to the degree of attention paid to the education experience part and the work experience part. For example, if more attention is paid to the education experience part, a larger weight will be assigned accordingly. Then, based on the assigned weights, the two first scores are weighted to obtain the final first score. Similarly, additional information can be extracted from the education experience part and the work experience part respectively. Therefore, a second score can be calculated for the education experience part, and another second score can be calculated for the work experience part. Then, in this case, the final second score can be calculated based on the two second scores for the education experience part and the work experience part. The specific calculation method can be to find the average of the two, or to assign different weights according to the degree of attention paid to the education experience part and the work experience part. For example, if more attention is paid to the education experience part, a larger weight will be assigned accordingly, and then the two second scores are weighted based on the assigned weights to obtain the final second score.

[0120] In another implementation, the education experience part and the work experience part are first integrated, and then the key information and experience information are extracted from the integrated content. Then, the first score can be calculated based on the extracted key information, and the second score can be calculated based on the extracted additional information. For example, after integrating the education experience part and the work experience part in the target resume document, the "education experience part: studied in a 211 college, a 2021 undergraduate computer science graduate, passed the English level 6 during undergraduate studies; work experience part: familiar with Python language, with software development project experience" is obtained. The integrated content is matched with the keywords in the key information database, and the matched keywords are: 211 college, undergraduate, computer science, and the matching score is 90, which is used as the first score. Then the integrated content is matched with the keywords in the additional information database, and the matched keywords are: English level 6, Python language, project experience, and the matching score is 80, which is used as the second score.

[0121] After obtaining the first score and the second score based on the education experience part and the work experience part, the experience similarity, position similarity, the first score and the second score are weighted to obtain the matching degree between the target resume document and the position description. If the matching degree reaches the preset threshold, it is determined that the target resume document and the position description are successfully matched; if the matching degree does not reach the preset threshold, it is determined that the target resume document and the position description are unmatched.

[0122] For example, a matching threshold is preset, and corresponding weights are assigned to the experience similarity, position similarity, key information conditions and additional information conditions according to actual needs. The text content of the work experience part in the target resume document and the text content of the job requirements part in the job description are input into the pre-trained BERT model to obtain the experience similarity score of the work experience part and the job requirements part; the text content of the job requirements part in the target resume document and the text content of the position part in the job description are input into the trained BERT model to obtain the job similarity score of the job requirements part and the position part; the text content of the education experience part and the work experience part are keyword matched with the keywords in the key information database in the server to obtain a first score; the text content of the education experience part and the work experience part are keyword matched with the keywords in the additional information database in the server to obtain a second score; then based on the weights corresponding to the experience similarity, position similarity, key information conditions and additional information conditions, the experience similarity score, position similarity score, first score and second score are weighted to obtain a weighted result. If the weight is greater than the preset matching threshold, it is determined that the target resume document is successfully matched with the job description.

[0123] In the above implementation, the text content of the work experience part and the text content of the job requirements part are input into the semantic similarity model to obtain the experience similarity score of the work experience part and the job requirements part; the text content of the job requirements part and the text content of the job part are input into the semantic similarity model to obtain the job similarity score of the job requirements part and the job part; then the key information and additional information are extracted from the education experience part and the work experience part, and the first score is assigned based on the matching result of the key information and the preset key information condition; the second score is assigned based on the matching result of the additional information and the preset additional information condition; finally, the experience similarity score, the job similarity score, the first score, and the second score are weighted to obtain the matching degree of the target resume document and the job description; according to the matching degree, the matching result of the target resume document and the job description is determined. Through the above method, the matching result of the target resume document and the job description is jointly determined based on the experience similarity, the job similarity, the first score, and the second score, so as to more accurately determine the semantic similarity of the target resume document and the job description, dig out the semantic information between the target resume document and the job description, and improve the accuracy of the matching result.

[0124] based on Figure 6 , Figure 7 In the embodiment shown, after the step of extracting the text content in the target resume document, the indicated target position can also be determined. Accordingly, the step of determining the target position is specifically as follows: Figure 8 As shown, Figure 8 A flow chart of a method for determining a target position according to an embodiment of the present application is shown, which specifically includes the following steps:

[0125] S802. Extract keywords from the work experience part and the job requirements part respectively.

[0126] S804: Take the intersection of the keywords extracted from the work experience part and the keywords extracted from the job requirements part.

[0127] S806. Determine the target position based on the intersection.

[0128] The target resume document includes work experience, education, and job requirements; the job description includes the job position and job requirements. Job requirements refer to the job that the applicant needs. Job requirements refer to the specific requirements of the employer for the job.

[0129] In the actual application of resume document matching, sometimes the applicant does not clearly describe the job requirements. For example, the applicant fills in the job requirements section of the resume document as a test engineer, while the positions provided by the employer include software test engineer and hardware test engineer. If the text content of the job requirements section and the text content of the position section are directly input into the semantic similarity model, the position matching error will occur. Therefore, before matching the target resume document, it is necessary to determine the target position based on the work experience section in the target resume document and the job requirements section in the job description.

[0130] In a possible implementation of an embodiment of the present application, after extracting the text content in the target resume document, the keywords of the work experience portion of the text content are extracted and compared with the keywords of the job requirements portion of the job description, the keyword intersection is extracted as the comparison result, and then the specific target position is determined based on the comparison result.

[0131] For example, the position requirement in the target resume document is "test engineer", and the work experience is "experienced in software testing projects, familiar with Java and other programming languages, participated in software development projects, and received professional training in software testing"; the position in the job description is "software testing engineer, hardware testing engineer", and the corresponding position requirement for software testing engineer is "familiar with programming languages ​​and software testing experience", and the corresponding position requirement for hardware testing engineer is "skilled in welding and repairing electronic components, mastering the hardware structure of hardware products, and familiar with the basic principles of circuits". The keywords extracted from the work experience part of the target resume document are "software testing, Java, programming language, software development, software testing", and the keywords extracted from the position requirements corresponding to the position part in the job description are "programming language, software testing" for software testing engineers, and "electronic components, hardware, circuits" for hardware testing engineers. The intersection of the keywords extracted from the work experience part of the target resume document and the position requirements part in the job description is "programming language, software testing", that is, the target position corresponding to the "test engineer" filled in the position requirement part of the target resume document is determined to be "software testing engineer".

[0132] In the above implementation, keywords are extracted from the work experience part and the job requirements part respectively, and the intersection of the keywords extracted from the work experience part and the keywords extracted from the job requirements part is taken, and the target position is determined according to the intersection. Through the above method, the target position corresponding to the job requirements in the target resume document in the job description can be determined, avoiding the situation where the target resume document is matched incorrectly due to unclear job requirements, and improving the accuracy of the target resume document matching.

[0133] based on Figure 2 In the embodiment shown, after the step of extracting the text content in the target resume document, a recommendation of the content to be added to the resume can also be made. Accordingly, the step of recommending the content to be added to the resume is as follows: Fig. 9 As shown, Fig. 9 A flow chart of a method for recommending content to be added to a resume according to an embodiment of the present application is shown, which specifically includes the following steps:

[0134] S902: Compare the text content with the job description to determine the resume content to be added that is included in the job description and not included in the target resume document.

[0135] S904: Send recommended adding information to the terminal, wherein the recommended adding information includes the content to be added to the resume.

[0136] After obtaining the target resume document to be matched and the preset job description, the text content in the target resume document is extracted, and then the text content is compared with the job description to determine the resume content to be added that is included in the job description and not included in the target resume document, and the resume content to be added is sent to the terminal as recommended addition information.

[0137] In one possible implementation method of the embodiment of the present application, after the text content of the target resume document is standardized according to the specified resume format, the result of the standardized processing is: "Basic information part: basic information, age, gender; Education experience part: graduation school, major courses, professional ability; Work experience part: professional experience"; then the text content in the job description is standardized according to the specified resume format, and the result of the standardized processing is: "Basic information part: basic information, age, gender; Education experience part: graduation school, major courses, professional ability; Work experience part: professional experience; Skills and expertise part: C language", and the text content in the target resume document is compared with the text content in the job description. The comparison result shows that the text content of the job description contains the skills and expertise part, while the text content of the target resume document does not contain the skills and expertise part. Therefore, the skills and expertise part in the job description is sent to the terminal as recommended additional information. Through the above method, the applicant can modify the resume document according to the recommended additional information, thereby improving the accuracy of the target resume document matching.

[0138] based on Figure 2 In the embodiment shown, after the step of obtaining the target resume document, the target resume document can also be modified. Accordingly, the step of modifying the target resume document is as follows: Fig.10 As shown, Fig.10 A flow chart of a method for modifying a target resume document provided in accordance with an embodiment of the present application is shown, which specifically includes the following steps:

[0139] S1002. Analyze the target resume document to determine defects in the target resume document.

[0140] S1004. Determine the correction strategy corresponding to the defect.

[0141] S1006: Send recommended correction information to the terminal, where the recommended correction information carries an identifier of the defect and a correction strategy corresponding to the defect.

[0142] The defect of the target resume document refers to the poor organization of the text content in the target resume document and / or the low similarity between the text content of the target resume document and the job description. The correction strategy refers to different correction tips taken for different defects. The recommended correction information carries the defect identifier and the correction strategy corresponding to the defect.

[0143] In this embodiment, after obtaining the target resume document to be matched, the target resume document will be analyzed to determine whether the target resume document has poor text content organization and whether the text content of the target resume document has low similarity with the job description. If it is determined that the target resume document has defects, the defective part will be marked and a corresponding correction prompt will be sent to the terminal based on the existing defects. Among them, marking the defective part can be making the defective part bold, underlining, adding a background color, etc.

[0144] In a first possible implementation method of an embodiment of the present application, after the target resume document to be matched is obtained, paragraph recognition is performed on the target resume document. For example, if there is a space in front of a line of text, the line of text is recognized as the beginning of a paragraph, and if there are multiple spaces after a line of text, it is recognized as the end of a paragraph. If it is recognized that the target resume document is only one paragraph, it is determined that the text content of the target resume document is poorly organized. Accordingly, the resume text content is bolded, and a correction message of "please write in paragraphs" is sent to the terminal to address the defect of poor organization of the text content.

[0145] In a second possible implementation method of the embodiment of the present application, a company profile database is created in advance in the recruiter's server or terminal, and the database may contain the names and profiles of companies across the country. After obtaining the target resume document to be matched, if it is identified that the target resume document contains a description of "company", the company description is matched with the company name in the company profile database by keywords. If the keyword match fails, it is determined that the defect of the target resume document is that the work experience description is unclear. Accordingly, an underline is added to the company description part in the target position description as a defect mark, and a correction message of "supplementing the company profile" is sent to the terminal.

[0146] In a third possible implementation method of the embodiment of the present application, after the target resume document to be matched is obtained, the semantic similarity between the target resume document and the job description is determined based on the text content of the target resume document and the job description. If the similarity is lower than a preset threshold, it is determined that the similarity between the target resume document and the job description is too low. Accordingly, a background color is added to the portion of the target resume document that has a low similarity with the job description as a defect marker and correction information of "supplementing the resume document" is sent to the terminal.

[0147] In the above implementation, after the step of obtaining the target resume document to be matched, the target resume document is analyzed to determine the defects in the target resume document; then the correction strategy corresponding to the defects is determined, and recommended correction information is sent to the terminal, thereby avoiding the situation where the target resume document is matched incorrectly due to defects in the target resume document, and improving the accuracy of the target resume document matching.

[0148] based on Figure 2 In the embodiment shown, after the step of extracting the text content in the target resume document, error sentence detection can also be performed. Accordingly, the step of error sentence detection is specifically as follows: Fig.11 As shown, Fig.11 A flowchart of a method for detecting an incorrect sentence according to an embodiment of the present application is shown, which specifically includes the following steps:

[0149] S1102: Detect wrong sentences in the text content using preset error detection rules.

[0150] S1104: If an error is detected in the text content, a prompt indicating the error is sent to the terminal.

[0151] Error detection rules refer to the rules for detecting incorrect sentences in the target resume document.

[0152] In a possible implementation of the embodiment of the present application, the wrong sentence detection model first segments the sentences in the target resume document through the stammering word segmenter. Since the sentences may contain typos, the wrong sentences are detected from the two aspects of character granularity and word granularity. Among them, the character granularity can detect the likelihood probability value of a certain word in the sentence through the language model perplexity detection. If it is lower than the average value of the sentence text, it is judged that the probability of the word is a suspected typo is high; the word granularity can segment the sentence through the stammering word segmenter, and then match the segmented words with the words in the preset dictionary library. If the match fails, it is judged that the word is likely to be a typo. If it is detected that there are typos or wrong words in the sentence of the text content, the sentence is judged to be an incorrect sentence, and an incorrect sentence prompt is sent to the terminal.

[0153] In another possible implementation of the embodiment of the present application, the BERT error detection model can be used to detect error sentences in the text content, wherein the BERT error detection model is composed of a similar word detection module, a repeated word detection module, a professional vocabulary detection module, and a detection module based on deep learning. After extracting the text content in the target resume document, the text content is input into the BERT error detection model for error sentence detection. If an error sentence is detected in the text content, an error sentence prompt is sent to the terminal.

[0154] Through the above method, the situation where the target document matching measures occur due to errors in the target resume document can be avoided, thereby improving the accuracy of the target resume document matching.

[0155] Fig.12 A flowchart of another resume document matching method provided according to an embodiment of the present application is shown.

[0156] Input the target resume document to be matched, standardize the target resume document to be matched and the preset job description respectively, and extract the text content of each part of the standardized target resume document and the job description. Input the text content of the work experience part of the target resume document and the text content of the job requirements part of the job description into the semantic similarity model to obtain experience similarity; input the text content of the job requirements part of the target resume document and the text content of the position part of the job description into the semantic similarity model to obtain position similarity; extract key information and additional information from the education experience part of the target resume document, and extract key information and additional information from the work experience part, match the key information of the education experience part and the work experience part with the preset key information conditions, and obtain the key information matching result; match the additional information of the education experience part and the work experience part with the preset additional information conditions, and obtain the additional information matching result; weight based on experience similarity, position similarity, key information matching result, and additional information matching result, and finally obtain the target resume document matching result.

[0157] After extracting the text content of the target resume document, a fine-grained job analysis will be performed based on the work experience section in the target resume document and the job requirements section in the job description to determine the target job. Additional information will also be identified for the target resume document, that is, the text content in the target resume document will be compared with the job description to determine the resume content to be added that is included in the job description but not in the target resume document, and recommended additional information will be generated based on the resume content to be added.

[0158] After obtaining the target resume document to be matched, the target resume document will be analyzed and defects will be identified, that is, the defects in the target document will be determined, and corresponding correction information will be generated based on the existing defects; the text content in the target resume document will also be identified for incorrect sentences using preset error detection rules, that is, if incorrect sentences are detected in the text content, an incorrect sentence prompt will be generated.

[0159] After determining the target position for the target resume document and generating recommended added information, correction information, and error prompts, the server will integrate the above results to generate resume suggestions for improving the target resume document and send the resume suggestions to the terminal.

[0160] Through the above method, the semantic similarity between the target resume document and the job description is determined based on the text content and job description extracted from the target resume document, the semantic information between the target resume document and the job description is mined, and resume document matching is performed based on the semantic similarity between the target resume document and the job description, thereby improving the accuracy of the matching results.

[0161] Corresponding to the above method embodiment, the present application also provides a resume document matching device embodiment, Fig.13 FIG. 1 is a schematic diagram showing the structure of a resume document matching device provided in an embodiment of the present application. Fig.13 As shown, the device comprises:

[0162] The acquisition module 1302 is configured to acquire a target resume document to be matched and a preset job description;

[0163] Extraction module 1304, configured to extract text content in the target resume document;

[0164] A determination module 1306 is configured to determine the semantic similarity between the target resume document and the job description based on the text content and the job description;

[0165] The matching module 1308 is configured to determine the matching result between the target resume document and the job description based on semantic similarity.

[0166] Optionally, the extraction module 1304 is further configured to perform standardization processing on the target resume document based on a specified resume format; and extract the text content of each part of the target resume document after the standardization processing.

[0167] Optionally, the device further comprises:

[0168] The job description processing module is configured to perform standardization processing on the job description based on a specified resume format; extract the text content of each part of the job description after the standardization processing;

[0169] Optionally, determination module 1306 is further configured to determine the semantic similarity between each part in the target resume document and the corresponding part in the job description based on the text content of each part in the target resume document and the text content of the corresponding part in the job description; and weight each semantic similarity to obtain the semantic similarity between the target resume document and the job description.

[0170] Optionally, the target resume document includes a work experience section; the job description includes a job requirements section;

[0171] Determination module 1306 is further configured to match the text content of the work experience part with the text content of the job requirements part to obtain the work experience overlap; if the work experience overlap is greater than a preset threshold, the text content of the work experience part and the text content of the job requirements part are input into a pre-trained semantic similarity model to obtain the experience similarity between the work experience part and the job requirements part.

[0172] Optionally, the target resume document further includes an education experience section and a job requirements section; the job description further includes a job position section;

[0173] The determination module 1306 is further configured to input the text content of the position requirement part and the text content of the position part into a semantic similarity model to obtain the position similarity between the position requirement part and the position part; extract key information and additional information from the education experience part and the work experience part; assign a first score based on the matching result of the key information and the preset key information condition; assign a second score based on the matching result of the additional information and the preset additional information condition;

[0174] The matching module 1308 is further configured to weight the experience similarity, the position similarity, the first score, and the second score to obtain the matching degree between the target resume document and the position description; and determine the matching result between the target resume document and the position description based on the matching degree.

[0175] Optionally, the device further comprises:

[0176] A keyword extraction module is configured to extract keywords from the work experience part and the job requirements part respectively;

[0177] An intersection module is configured to take the intersection of the keywords extracted from the work experience part and the keywords extracted from the job requirements part;

[0178] The target position determination module is configured to determine the target position according to the intersection.

[0179] Optionally, the device further comprises:

[0180] A comparison module is configured to compare the text content with the job description to determine the resume content to be added that is included in the job description and not included in the target resume document;

[0181] The recommended addition information sending module is configured to send the recommended addition information to the terminal, wherein the recommended addition information includes the content to be added to the resume.

[0182] Optionally, the device further comprises:

[0183] An analysis module is configured to analyze the target resume document to determine defects in the target resume document;

[0184] A correction strategy module, configured to determine a correction strategy corresponding to the defect;

[0185] The recommended correction information sending module is configured to send the recommended correction information to the terminal, wherein the recommended correction information carries the identifier of the defect and the correction strategy corresponding to the defect.

[0186] Optionally, the device further comprises:

[0187] The detection module is configured to detect incorrect sentences in the text content using a preset error detection rule;

[0188] The error sentence prompt sending module is configured to send an error sentence prompt to the terminal if an error sentence is detected in the text content.

[0189] The resume document matching device provided by the present application obtains the target resume document to be matched and the preset job description, then extracts the text content in the target resume document, and determines the semantic similarity between the target resume document and the job description based on the extracted text content and the job description. Based on the semantic similarity, the matching result of the target resume document and the job description can be determined. Through the above method, the semantic similarity between the target resume document and the job description is determined based on the text content and the job description extracted from the target resume document, the semantic information between the target resume document and the job description is mined, and the resume document is matched based on the semantic similarity between the target resume document and the job description, thereby improving the accuracy of the matching result.

[0190] The above is a schematic scheme of a resume document matching device of the present embodiment. It should be noted that the technical scheme of the resume document matching device and the technical scheme of the resume document matching method mentioned above belong to the same concept. For the details not described in detail in the technical scheme of the resume document matching device, please refer to the description of the technical scheme of the resume document matching method mentioned above. In addition, each component in the device embodiment should be understood as a functional module that must be established to implement each step of the program flow or each step of the method, and each functional module is not an actual functional division or separation definition. The device claim defined by such a group of functional modules should be understood as a functional module architecture that mainly implements the solution through the computer program recorded in the specification, and should not be understood as a physical device that mainly implements the solution through hardware.

[0191] Fig.14 The structure block diagram of a computing device 1400 provided according to an embodiment of the present application is shown. The components of the computing device 1400 include but are not limited to a memory 1410 and a processor 1420. The processor 1420 is connected to the memory 1410 via a bus 1430, and the database 1450 is used to store data.

[0192] The computing device 1400 also includes an access device 1440 that enables the computing device 1400 to communicate via one or more networks 1460. Examples of these networks include a public switched telephone network (PSTN), a local area network (LAN), a wide area network (WAN), a personal area network (PAN), or a combination of communication networks such as the Internet. The access device 1440 may include one or more of any type of network interface (e.g., a network interface card (NIC)) whether wired or wireless, such as an IEEE 802.11 wireless local area network (WLAN) wireless interface, a World Wide Interoperability for Microwave Access (Wi-MAX) interface, an Ethernet interface, a universal serial bus (USB) interface, a cellular network interface, a Bluetooth interface, a near field communication (NFC) interface, and the like.

[0193] In one embodiment of the present application, the above components of the computing device 1400 and Fig.14 Other components not shown in the figure may also be connected to each other, for example, via a bus. It should be understood that Fig.14 The computing device structure block diagram shown is only for the purpose of illustration, and is not intended to limit the scope of the present application. Those skilled in the art may add or replace other components as needed.

[0194] Computing device 1400 may be any type of stationary or mobile computing device, including a mobile computer or mobile computing device (e.g., a tablet computer, a personal digital assistant, a laptop computer, a notebook computer, a netbook, etc.), a mobile phone (e.g., a smart phone), a wearable computing device (e.g., a smart watch, smart glasses, etc.), or other types of mobile devices, or a stationary computing device such as a desktop computer or PC. Computing device 1400 may also be a mobile or stationary server.

[0195] Among them, the processor 1420 is used to execute computer executable instructions of the resume document matching method.

[0196] The above is a schematic scheme of a computing device of this embodiment. It should be noted that the technical scheme of the computing device and the technical scheme of the resume document matching method described above are of the same concept, and the details not described in detail in the technical scheme of the computing device can be found in the description of the technical scheme of the resume document matching method described above.

[0197] An embodiment of the present application also provides a computer-readable storage medium storing computer instructions, which are used in a resume document matching method when executed by a processor.

[0198] The above is a schematic scheme of a computer-readable storage medium of this embodiment. It should be noted that the technical scheme of the storage medium and the technical scheme of the resume document matching method described above are of the same concept, and the details not described in detail in the technical scheme of the storage medium can be found in the description of the technical scheme of the resume document matching method described above.

[0199] The above describes specific embodiments of the present application. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recorded in the claims can be performed in an order different from that in the embodiments and still achieve the desired results. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0200] The computer instructions include computer program codes, which may be in source code form, object code form, executable files or some intermediate forms, etc. The computer readable medium may include: any entity or device capable of carrying the computer program code, recording medium, USB flash drive, mobile hard disk, magnetic disk, optical disk, computer memory, read-only memory (ROM), random access memory (RAM), electric carrier signal, telecommunication signal and software distribution medium, etc.

[0201] It should be noted that, for the above-mentioned method embodiments, for the sake of simplicity of description, they are all expressed as a series of action combinations, but those skilled in the art should be aware that the present application is not limited by the described action sequence, because according to the present application, certain steps can be performed in other sequences or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions and modules involved are not necessarily required by the present application.

[0202] In the above embodiments, the description of each embodiment has its own emphasis. For parts that are not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0203] The preferred embodiments of the present application disclosed above are only used to help explain the present application. The optional embodiments do not describe all the details in detail, nor do they limit the invention to the specific implementation methods described. Obviously, many modifications and changes can be made according to the content of the present application. The present application selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present application, so that those skilled in the art can understand and use the present application well. The present application is only limited by the claims and their full scope and equivalents.

Claims

1. A resume document matching method, characterized in that: include: Obtain the target resume document to be matched and the preset job description; Extracting text content from the target resume document; Determining the semantic similarity between the target resume document and the job description according to the text content and the job description, the target resume document includes a work experience section, an education experience section and a job requirement section, and the job description includes a job requirement section and a job position section, and determining the semantic similarity between the target resume document and the job description according to the text content and the job description includes: inputting keywords of the work experience section and the job requirement section into a pre-trained semantic similarity model to obtain the experience similarity between the work experience section and the job requirement section; performing a similarity comparison between the job requirement section and the job section to obtain the job similarity; extracting key information and additional information from the education experience section and the work experience section, respectively, the key information being the basic requirements of the recruiter for the applicant, and the additional information being the bonus items preset by the recruiter for the applicant; allocating a first score based on the matching result of the key information and the preset key information condition; and allocating a second score based on the matching result of the additional information and the preset additional information condition; Based on the semantic similarity, determining the matching result between the target resume document and the job description, including: weighting the experience similarity, the job similarity, the first score, and the second score to obtain the matching degree between the target resume document and the job description; determining the matching result between the target resume document and the job description based on the matching degree.

2. The method according to claim 1, characterized in that The step of extracting text content from the target resume document comprises: Based on the specified resume format, standardize the target resume document; Extract the text content of each part of the target resume document after standardization.

3. The method according to claim 1, characterized in that Before the step of determining the semantic similarity between the target resume document and the job description based on the text content and the job description, the method further includes: Standardize the job description based on the specified resume format; Extract the text content of each part of the job description after standardization.

4. The method according to any one of claims 1 to 3, characterized in that The step of determining the semantic similarity between the target resume document and the job description based on the text content and the job description includes: Determine the semantic similarity between each part in the target resume document and the corresponding part in the job description according to the text content of each part in the target resume document and the text content of the corresponding part in the job description; Each of the semantic similarities is weighted to obtain the semantic similarity between the target resume document and the job description.

5. According to the method of claim 1, the step of matching the text content of the work experience part with the text content of the job requirements part to obtain the experience similarity comprises: Matching the text content of the work experience part with the text content of the job requirements part to obtain the degree of overlap of work experience; If the work experience overlap is greater than a preset threshold, the text content of the work experience part and the text content of the job requirements part are input into a pre-trained semantic similarity model to obtain the experience similarity between the work experience part and the job requirements part.

6. The method according to claim 5, characterized in that After the step of extracting the text content in the target resume document, the method further includes: Extract keywords from the work experience part and the job requirements part respectively; Taking the intersection of the keywords extracted from the work experience part and the keywords extracted from the job requirements part; Based on the intersection, the target position is determined.

7. The method according to any one of claims 1 to 3, characterized in that After the step of extracting the text content in the target resume document, the method further includes: Comparing the text content with the job description to determine the resume content to be added that is included in the job description and not included in the target resume document; Sending recommended adding information to the terminal, wherein the recommended adding information includes the content to be added to the resume.

8. The method according to any one of claims 1 to 3, characterized in that After the step of obtaining the target resume document to be matched, the method further includes: Analyze the target resume document to determine defects in the target resume document; Determining a correction strategy corresponding to the defect; Sending recommended correction information to the terminal, the recommended correction information carrying an identifier of the defect and a correction strategy corresponding to the defect.

9. The method according to any one of claims 1 to 3, characterized in that: After the step of extracting the text content in the target resume document, the method further includes: Using preset error detection rules to detect incorrect sentences in the text content; If it is detected that there is an error in the text content, an error prompt is sent to the terminal.

10. A resume document matching device, characterized in that: include: An acquisition module is configured to acquire a target resume document to be matched and a preset job description; An extraction module, configured to extract text content from the target resume document; A determination module is configured to determine the semantic similarity between the target resume document and the job description according to the text content and the job description, the target resume document includes a work experience part, an education experience part and a job requirement part, and the job description includes a job requirement part and a job position part, and the determining the semantic similarity between the target resume document and the job description according to the text content and the job description includes: inputting keywords of the work experience part and the job requirement part into a pre-trained semantic similarity model to obtain the experience similarity between the work experience part and the job requirement part; performing a similarity comparison between the job requirement part and the job part to obtain the job similarity; extracting key information and additional information from the education experience part and the work experience part, respectively, the key information being the basic requirements of the recruiter for the applicant, and the additional information being the bonus items preset by the recruiter for the applicant; allocating a first score based on the matching result of the key information and the preset key information condition; and allocating a second score based on the matching result of the additional information and the preset additional information condition; A matching module is configured to determine the matching result between the target resume document and the job description based on the semantic similarity, including: weighting the experience similarity, the job similarity, the first score, and the second score to obtain the matching degree between the target resume document and the job description; and determining the matching result between the target resume document and the job description based on the matching degree.

11. A computing device, characterized in that: include: Memory and processor; The memory is used to store computer executable instructions, and the processor is used to execute the computer executable instructions to implement the steps of the resume document matching method described in any one of claims 1 to 9.

12. A computer-readable storage medium storing computer instructions, characterized in that: When the instruction is executed by the processor, the steps of the resume document matching method described in any one of claims 1 to 9 are implemented.

Images