Recruitment content generation method and generation device

Abstract

The present application discloses a method and device for generating recruitment content. The method comprises: receiving a job description from an HR client; processing the job description by identifying a set of target attributes, wherein the set of target attributes comprises a formatted job requirement, a keyword, a classification, and an associated list; generating a job announcement based on a job description template; generating a standardized job description; generating a personalized job description based on talent preferences; generating personalized job description content based on TS; and pushing the generated personalized job description content to TS.

Description

[0001] This application is a divisional application of the Chinese invention patent application with the application date of October 16, 2020, application number 2020111127887, and name “Overall recruitment system based on talent discoverer”. Technical Field

[0002] The present invention generally relates to recruitment systems, and more particularly, to a Talent Scouting (TS) based recruitment system. Background Art

[0003] The explosive growth of social networks has provided a wide range of benefits to people. Members of social networks share a wide variety of information through social media platforms. People rely on social networks (also known as social media networks) to get more connections and get information faster. In recent years, social media networks have expanded beyond informal networking between friends and family. Social media platforms have been increasingly used for business and professional purposes. Many organizations and business units have their social media presence. Social media marketing has gained tremendous momentum. The rapid and massive adoption of social media technologies by businesses and professionals has changed how businesses find partners, how people access information, and how people keep up with the latest developments in their professional fields. In one notable area, social media has been increasingly seen as a professional platform to help businesses and job seekers connect. Businesses with job vacancies post their job descriptions on social media networks. Job seekers or professionals post their resumes and interests on social media networks. Social media networks provide users with connections to each other and participation in groups. Although current social media provides advantages for businesses and talents to be connected, the recruitment process is still inefficient. With the explosive growth of information on social media networks, a lot of effort is needed to find the right match quickly and effectively. Some improvements were developed, injected into the candidate ranking system, but it still could not keep up with the exponential growth of information flow on social networks.

[0004] Some improvements are needed to use social media for recruiting more effectively. Summary of the invention

[0005] Provided is a method and system for a recruitment system based on a talent scout (TS). In a novel aspect, a group of talent scouts is identified for the recruitment of target talents, and personalized recruitment content is generated based on the job description and the attributes of the corresponding TS. A group of TSs is identified from a social network, wherein each identified TS is assigned a set of TS attributes, and wherein each TS has a social media influence value higher than a predefined influence threshold. A job recruitment request is processed by assigning a set of target attributes based on the recruitment request. A subgroup of TSs is selected from the group of identified TSs by comparing the set of TS attributes of each TS in the TS group with the set of target attributes based on one or more predefined matching criteria. At least based on the set of target attributes and the set of TS attributes of each corresponding TS, TS recruitment content is created for each interested TS in the subgroup of TS. The created recruitment content is sent to each corresponding interested TS in the subgroup of the selected TS, wherein a recruitment announcement based on the recruitment content is published to the followers of each corresponding interested TS. In one embodiment, the TS attributes include one or more TS elements, and the one or more TS elements include: a multi-level industry classification field, a social media influence value, a follower field, applicable keywords, a feedback field, and a last updated timestamp. In another embodiment, the target attribute includes one or more target elements, and the one or more target elements include: a multi-level industry classification field, a qualification list, applicable keywords, preferences of target talents, and associated fields. In one embodiment, identifying a group of TS involves: collecting information from a social network, and processing the collected information using an artificial neural network (ANN) implemented in a computer system to assign a set of TS attributes. In another embodiment, the recruitment request is an unformatted input, and wherein processing the recruitment request involves: processing the received recruitment request using a recurrent neural network (RNN) implemented in a computer system. In yet another embodiment, the TS recruitment content is based on previously generated standardized recruitment content, and wherein the standardized recruitment content is based on a predefined recruitment template. In one embodiment, the recruitment content includes one or more personalized content, and the one or more personalized content includes: personalized job descriptions, industry news, training information, and professional service information. In another embodiment, the personalized content is generated using artificial intelligence (AI) deep learning implemented in a computer system based on at least a set of target attributes and a corresponding set of TS attributes. In one embodiment, a set of behavioral data associated with each recruitment announcement is obtained, wherein the set of behavioral data includes one or more elements, and the one or more elements include: the number of readings, the number of clicks on the links in the announcement, the number of responses, and the number of forwardings. The collection of behavioral data is processed using machine learning implemented in a computer system to generate a talent profile.In another embodiment, each corresponding TS recruitment content is updated based on the generated talent profile using deep learning implemented in the computer system. In yet another embodiment, one or more responses to job openings in corresponding recruitment notices published by the TS are received and analyzed. Recommendations are made based on at least a set of corresponding applicant attributes and a set of target attributes in accordance with one or more predefined rules.

[0006] Other embodiments and advantages are described in the following detailed description. This summary is not intended to limit the invention. The invention is defined by the claims. BRIEF DESCRIPTION OF THE DRAWINGS

[0007] The accompanying drawings illustrate embodiments of the present invention, wherein like reference numerals denote like components.

[0008] Figure 1 An exemplary diagram showing a Talent Scout (TS)-based recruitment system according to an embodiment of the present invention is shown.

[0009] Figure 2 An exemplary diagram of a TS-based intelligent overall recruitment system having an HR client, a talent processing computer system, talent prospectors, and talents according to an embodiment of the present invention is shown.

[0010] Figure 3 An exemplary diagram showing a TS-based recruitment control device according to an embodiment of the present invention.

[0011] Figure 4 An exemplary diagram showing identification of a TS according to an embodiment of the present invention is shown.

[0012] Figure 5 An exemplary diagram showing the use of a TS-based recruitment system for personalized recruitment content according to an embodiment of the present invention.

[0013] Figure 6 An exemplary diagram from a talent management system through a TS-based recruitment system according to an embodiment of the present invention is shown.

[0014] Figure 7 An exemplary flow chart of a TS-based recruitment system according to an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0015] Reference will now be made in detail to some embodiments of the present invention, examples of which are illustrated in the accompanying drawings.

[0016] Social media platforms provide members to establish or join online communities, obtain information, interact with each other, and share information. With the exponential growth of social networks, social media influencers (e.g., Internet celebrities) are emerging. Social media influencers have gained and established credibility in one or more special industry classifications through his / her presence on social media platforms. When members of social media networks share information with the public or one or more online communities, the group of members builds their reputation and gains trust in their knowledge and expertise in one or more industry classifications. These members increase the number of their followers on social media networks and use their influence through social media. Social media influencers can obtain a large audience in one or more target industry classifications by establishing a large network of followers. In a novel aspect, a recruitment system based on talent scouts is provided. Talent scouts are social media influencers who are recruited to approach target talent groups for recruitment. TS-based recruitment systems provide large, targeted talent pools for recruitment businesses. In addition, TS-based recruitment systems provide credible and relevant career market information for talents. In addition, TS-based recruitment systems provide a two-way feedback mechanism to improve the efficiency of talent search.

[0017] Figure 1An exemplary diagram of a recruitment system based on a talent scout (TS) according to an embodiment of the present invention is shown. The exemplary TS-based recruitment system 100 includes a talent processing computer system 101, a talent scout group subsystem 110, a human resources (HR) client subsystem 120, a talent subsystem 130, a social media platform 140, and a blockchain database 150. In one novel aspect, one or more talent scouts, such as a talent scout group 110, are identified. The TS group 110 receives recruitment content and employment decisions and feedback on talents from the talent processing computer system 101, and sends resumes and other behavioral data to the talent processing computer system 101. The TS group 110 communicates with the talent group 130 via the social media platform 140 by posting job-related content to followers in the talent group 130 and receiving resumes from interested followers of the talent group 130. The HR client 120 generates job requirements and employment decisions and feeds back to the talent processing computer system 101. The HR client 120 receives candidate information from the talent processing computer system 101. The candidate information is received from members of the identified TS group 110. In one embodiment, the talent processing computer system also performs automated, AI-enabled testing and / or interviews on interested talents from the talent group 130 based on the job requirements received from members of the HR client 120. The talent processing computer system 101 processes the application upon receiving an application and / or resume from a talent in the talent group 130. In one embodiment, AI-based machine learning is implemented to see if the application matches the job requirements. If a match is found, an automatic online test is performed by the talent processing computer system 101. In one embodiment, in addition to one or more written tests performed, a chatbot interview is also given. The talent processing computer system 101 sends the test / interview results to interested talents from the talent group 130. In one embodiment, the test / interview results include at least one of the following: the correctness of the answer, the sentiment analysis of the chatbot interview, and the clarity evaluation of the voice and expression based on the chatbot interview. The talent processing computer system 101 also sends employment decisions and / or feedback to interested talents in the talent group 130. In one embodiment, all transaction records and documents are maintained in the blockchain database 150 and the Interplanetary File System (IPFS).

[0018] In one embodiment, a group of talent scouts 110 are identified from a larger group of talent scout candidates 119. Talent scouts are identified by selecting a group of members of social media that are social media influencers in one or more identified industry fields. In one embodiment, a group of members of social media are identified as talent scout candidates 119. Talent scouts are social media influencers. Social media influencers are identified based on predefined influencer identification rules.

[0019] In one embodiment, once a TS is identified and added to a TS-based recruitment system, the talent processing computer system 101 pushes recruitment content to the TS, including job announcements, articles, images and / or videos, relevant industry news, and career or professional training programs related to the TS's area of ​​interest. Upon receiving the recruitment content, the TS of the TS group 110 pushes the content to its followers who are members of the talent 130 through the social media channels of its field. In one embodiment, if the followers of the TS are interested in one or more job announcements, the followers can apply for the vacant positions and submit their resumes through the links provided by the TS. In one embodiment, the links are embedded in the application received by the followers.

[0020] Figure 2An exemplary diagram of an intelligent overall recruitment system based on TS with an HR client, a talent processing computer system, a talent prospector and a talent according to an embodiment of the present invention is shown. The intelligent overall recruitment system 200 includes an HR client subsystem 210, a talent processing computer system 220, a talent prospector subsystem 250 and a talent subsystem 260. The HR subsystem 210 includes a plurality of processing and / or functional modules, such as a job vacancy requirement module 211, a talent package processing module 212, a talent interview module 213, and an employment decision and feedback module 214. The talent processing computer system 220 includes one or more modules that interact with the HR client 210, including a client management and job description module 221, a TS identification module 222, a talent recommendation module 223 and a defect analysis module 224. The talent processing computer system 220 also includes internal modules, such as an AI-enabled personalized TS content creator 231, and an online test and chatbot Q&A management module 232. The talent processing computer system 220 also includes modules that interact with talent prospectors, such as a push package container module 241, a machine learning-based creator for talent profiles 242, an application processing module 243, and a recommendation generator 244. The talent prospector subsystem 250 includes: a process 251 that dynamically pushes content from a container for each follower; a behavioral data collector 252; application verification, suggestions and submission 253; and hiring decision, defect reporting and recommendation processing 254. The talent subsystem 260 includes a reading, purchase, subscription, application and forwarding collector 261, a job application processing 262, an online test and chatbot interview 263, and a recommendation and feedback receiver 264. Each subsystem includes one or more components such as software, hardware, and firmware to implement its functions. It can run on a single device or on multiple devices. Each module of the subsystem can be implemented with software, hardware, firmware, and a combination of the above. Each subsystem has: at least one processor coupled to at least one memory unit; one or more storage devices; and optional communication hardware and software for communicating with the subsystem inside and outside.

[0021] In one embodiment, client management 521 obtains job opening requirements from 511 and identifies TS based on the job opening requirements. The information of the identified TS and the feedback information of the talent behavior profile are inputs to the AI-enabled personalized content creator 231. The talent behavior profile is generated by machine learning, using data collected by the talent scout from his / her followers. Module 231 generates personalized recruitment content based on job requirements and pushes the package to the corresponding TS through container 241. When a talent applies for a job through a recruitment information posted by a TS, the application is processed by the TS subsystem 250 using a module 253 for verification and submission. The application is processed in the application processor 243. If 243 determines that the application meets the threshold requirements, the talent processing computer system 220 creates an online test and / or a chatbot Q&A. Applicants of the talent subsystem take online tests and / or participate in chatbot interviews. The results are passed to the HR client 210. When a hiring decision is made, the HR client subsystem notifies the talent processing computer system 220. The talent computer system 220 generates recommendations and / or feedback and sends them to the TS. Applicants to the talent subsystem receive hiring decisions and / or recommendations from the TS subsystem 250 .

[0022] Figure 3An exemplary diagram of a TS-based recruitment control device according to an embodiment of the present invention is shown. The TS-based recruitment control device 300 has one or more user and / or network interfaces 304 for sending and receiving one or more commands or signals from a user or a connected network, and at least one processor 301 coupled to at least one memory 302. The one or more processors 301 process the received user and / or network commands and signals, and call different functional modules to perform features in the TS-based recruitment control device. The memory 302 stores program instructions and data 303 to control the operation of the device 300. The control device 300 also includes control modules such as a client management module 310, a TS management module 320, a talent management module 330, and a recommendation engine 340. The client management module 310 performs a functional interface with a client subsystem such as an HR client. The client management module 310 also includes a job description generator module 311 and an employment decision and feedback module 312. The job description module 311 receives the original job description from the client and generates a personalized job description for a group of TS. The hiring decision and feedback module 312 receives hiring decisions and feedback from the client and notifies the TS and / or applicants. The TS management module 320 communicates with one or more TS subsystems and performs functions for the recruitment process. The TS management module also includes a TS identification module 321 and a recruitment content generator module 322. The TS identification module 321 identifies one or more TSs from the social media network and recruits the identified one or more TSs. The recruitment content generator generates recruitment-related content based on the job description and the attributes of the TS. The talent management module 330 manages talent-related functions. The talent management module 330 also includes: a talent-TS management module 331, which manages the interaction between one or more talents and one or more TSs; and a talent evaluation module 332, which performs talent evaluation based on information received from the talents. The above-mentioned functional modules can be implemented in software, hardware, firmware or a combination thereof.

[0023] In one novel aspect, one or more social media influencers are recruited as talent scouts. Talent scouts are social media influencers in one or more matching industry categories. A TS-based recruitment system obtains information of a group of talent scout candidates from a social media network, and identifies talent scouts based on the attributes of each talent scout candidate obtained and a set of identification rules. In one embodiment, AI-based machine learning such as an artificial neural network (ANN) is implemented to identify talent scouts.

[0024] Figure 4An exemplary diagram of identifying TS according to an embodiment of the present invention is shown. A social media network 401 has multiple talent scout candidates, each with multiple followers. Talent scout candidates publish content to their followers. In step 411, a TS-based recruitment system obtains TS candidate information from the social media network 401. In one embodiment, in step 413, a crawler is executed to collect information about TS candidates. The crawler starts with a list of member pages of the social media network. When the crawler visits these pages, it identifies all friend pages embedded in the current page and follows the identified pages. In one embodiment, the TS-based recruitment system also identifies a set of predefined TS attributes for each visited page. In another embodiment, the obtained TS candidate list is updated when one or more update criteria are met. In step 412, the obtained TS candidate information is processed to identify a group of talent scouts. Talent scouts are social media influencers and match one or more industry classifications. In one embodiment, an AI-based recognition program 416 such as an ANN is used based on the TS candidate information. In other embodiments, the TS candidate information is parsed into TS candidate attributes based on attributes and selection rules 415. Group talent scouts are identified using a combination of predefined selection rules and optional corresponding thresholds.

[0025] In one embodiment, social media influencers are identified by the number of followers being greater than a predefined influence threshold. In other embodiments, the number of corresponding published reads or clicks is compared with a predefined influence threshold to determine whether a member is a social media influencer. In other embodiments, a weighted value of influence is used. For example, the influence value is a weighted combination of the number of followers, the number of reads, and the number of clicks. In other embodiments, a combination of different thresholds and different attributes may also be used to identify social media influencers. In one embodiment, influencer identification rules may be dynamically updated. In other embodiments, influencer identification rules and / or attribute thresholds are based on industry categories. A subset of identified social media influencers is identified as talent scouts based at least on the industry category to which the talent scout candidate belongs. In one embodiment, members of a talent scout candidate group are identified as being in one or more industry categories / classifications. In other embodiments, the industry category of a TS candidate is based on one or more factors including the profile of the follower, the content of the post, and feedback. Rules and thresholds may be industry-specific and / or client-specific.

[0026] At step 421, the TS-based recruitment system determines whether the TS candidate is qualified as a talent scout. If step 421 determines that it is, the talent scout candidate is recruited as a talent scout. If step 421 determines that it is not, the TS candidate is discarded from the list. In one embodiment, the talent scout list is updated based on one or more update criteria. In one embodiment, at step 435, the recruited talent scouts and their corresponding TS attributes are stored in the TS database.

[0027] In a novel aspect, the TS-based recruitment system pushes personalized content to target talents through talent scouts based on one or more job descriptions from the client. In addition to personalized job descriptions for job vacancies, personalized content also includes personalized related content, such as industry news, professional training courses, and other relevant social or professional announcements.

[0028] Figure 5 An exemplary diagram of using a TS-based recruitment system for personalized recruitment content according to an embodiment of the present invention is shown. The TS-based recruitment system receives a recruitment request from an HR client 502, and selects a TS 501 based on the recruitment request and a set of TS attributes. TS 501 receives personalized recruitment content from the TS-based recruitment system and publishes 580 to its followers including target talents 503. In one embodiment, the job description 511 from the HR client 502 is in the original format, not in the TS-based format. In step 513, the TS-based recruitment system processes the job description 511 by identifying a set of target attributes, such as formatted job requirements, keywords, classifications, and associated lists. In one embodiment, step 513 uses an AI-based program to process the job description 511, such as a natural language processing (NLP) recurrent neural network (RNN) / natural language understanding (NLU) RNN. In another embodiment, HR 502 uses a formatted job description form at step 512 to generate a job description. The formatted job description generated at step 512 includes the target attributes required in the job description. At step 514, a job posting is generated based on the job description template. In one embodiment, the job template 561 has one or more job templates based on different industry categories. In other embodiments, the job template 561 is also based on other factors such as TS profiles. At step 515, a standardized job description is generated. At step 516, a personalized job description is generated based on talent preferences 562. At step 517, personalized job description content based on the TS is generated.

[0029] In one embodiment, in addition to the personalized job description, other personalized content is generated for the corresponding talent scout to publish. An exemplary personalized recruitment publication 580 includes personalized industry news 588, personalized training announcements 585, and personalized job descriptions 581. Other personalized content 586, such as personalized professional or entertainment announcements, may also be included. In one embodiment, personalized content such as those in 588, 585, 581, and 586 are generated by an AI-based implementation in a TS-based recruitment system. The AI-based implementation generates personalized content based on one or more attributes such as target attributes and TS attributes of the job description. Upon receiving these personalized contents, TS 501 publishes personalized content 580 to its followers. In one embodiment, talent 503 receives these contents through a mini-app. In another embodiment, the personalized content has an embedded link to additional information or applications when talent 503 clicks on it.

[0030] In one embodiment, the TS-based job description 517 is further evaluated at step 530 before being pushed to TS 501. Step 530 involves a client approval process 531, in which input is received from the HR client 502. The input includes one or more elements, such as approval or disapproval, updated content, and additional content. In one embodiment, talent feedback 532 received from TS 501 is also processed. Talent feedback includes one or more talent information, such as updated talent profiles. In step 521, the TS-based recruitment system determines whether the job description 517 needs to be updated. If step 521 determines no, the job description is used to generate recruitment content and pushed to TS 501 in step 551. If step 521 determines "no", the job description is updated accordingly at step 552.

[0031] Once the selected one or more TSs receive the personalized recruitment content, the content is published to their corresponding followers. When the target talent, who is a follower or member of the TS, receives the publication of the TS, the target talent can interact with a TS-based recruitment system such as a talent processing system through the TS or an embedded application. The target talent's application and / or resume is received by the talent processing system and processed to determine whether a match is found. The decision and / or feedback is communicated to the target talent.

[0032] Figure 6An exemplary diagram from a talent management system through a TS-based recruitment system according to an embodiment of the present invention is shown. In one embodiment, recruitment content is published using a mini-application. The talent processing system 650 has a mini-application container generator 651, an application processing unit 652, and a decision, feedback and recommendation application 654. The talent prospector 630 includes at least a personalized mini-application 631 for talents, a talent profile unit 632, an update auxiliary application 633 for qualified talents, and a talent notification unit 634 in the mini-application. The talent 610 includes at least a receiver 611 for personalized mini-applications, a machine learning and / or data mining application 615 for obtaining talent behavior data, and a response unit 616 for feedback and recommendation. The talent processing system 650 generates a mini-application container with personalized recruitment content and sends it to the corresponding TS 630. The TS 630 pushes the personalized recruitment content to the talent 610 through the mini-application. In one embodiment, before pushing the content to the talent 610 through the mini-application, the TS 630 further personalizes the content based on the talent profile 632. In one embodiment, the talent profile 632 is updated based on the behavioral data collected by the talent 610. In one embodiment, the behavioral data includes viewing history, click history, and other relevant activities of the talent 610. In one embodiment, the behavioral data is collected using machine learning and / or data mining programs. Upon receiving the mini-application, the talent 610 can view the content at 612 and send a resume at 613 if the talent 610 is interested. The TS 630, upon receiving the response and / or resume from the talent 610, processes the information at 633 and determines whether there is a match. In one embodiment, if the TS 630 determines that the talent 610 is qualified, the TS 630 sends an application update suggestion to the talent 610. The TS 630 forwards the application from the talent 610 to the talent processing system 650. The talent processing system 650 sends a hiring decision and / or feedback and recommendations to the talent 610 via the TS 630.

[0033] Figure 7An exemplary flow chart of a TS-based recruitment system according to an embodiment of the present invention is shown. At step 701, a computer system identifies a group of TSs from a social network, wherein each identified TS is assigned a set of TS attributes, and wherein each TS has a social media influence value higher than a predefined influence threshold. At step 702, the computer system processes a recruitment request by assigning a set of target attributes based on the recruitment request. At step 703, the computer system selects a subgroup of TSs from the identified group of TSs by comparing a set of TS attributes of each TS in the TS group with a set of target attributes based on one or more predefined matching criteria. At step 704, the computer system creates TS recruitment content for each interested TS in the subgroup of TSs based at least on the set of target attributes and the set of TS attributes of each corresponding TS. At step 705, the computer system sends the created recruitment content to each corresponding interested TS in the subgroup of the selected TS, wherein a recruitment announcement based on the recruitment content is published to followers of each corresponding interested TS.

[0034] Although the present invention has been described in conjunction with certain specific embodiments for instructional purposes, the invention is not limited thereto. Therefore, various modifications, adaptations and combinations of the various features of the described embodiments may be practiced without departing from the scope of the invention as set forth in the claims.

Claims

1. A method for generating recruitment content, characterized in that: The generation method includes: Receive job descriptions from HR clients; processing the job description by identifying a set of target attributes, wherein the set of target attributes includes a formatted list of job requirements, keywords, categories, and associations; Generate job announcements based on job description templates; Generate standardized job descriptions; Generate personalized job descriptions based on talent preferences; Generate personalized job description content based on TS; Push the generated personalized job description content to TS.

2. The generation method according to claim 1, characterized in that: Before pushing the generated personalized job description content to the TS, the generation method further includes: receiving input from the HR client; Based on the received input, determining whether the generated personalized job description content needs to be updated; If it is determined that an update is needed, the generated personalized job description content is updated.

3. The generation method according to claim 1 or 2, characterized in that: The generation method also includes: A TS is selected based on a recruitment request received from the HR client and a set of TS attributes.

4. The generation method according to claim 3, characterized in that: The selecting a TS based on the recruitment request received from the HR client and a set of TS attributes comprises: identifying a group of talent scouts TS from a social network, wherein each identified TS is assigned a set of TS attributes, and wherein each TS has a social media influence value above a predefined influence threshold; processing the job request by assigning a set of target attributes based on the job request; A subgroup of TSs is selected from the identified group of TSs by comparing the set of TS attributes of each TS in the group of TSs with the set of target attributes based on one or more predefined matching criteria.

5. The generation method according to claim 4, characterized in that: Identifying TS groups involves: executing a crawler program starting from a list of member pages of said social network to collect collected information of TS candidates; processing the collected information using an artificial neural network (ANN) implemented in the computer system to assign the set of TS attributes; as well as Groups of TSs are identified using a combination of predefined selection rules and optional corresponding thresholds, talent behavior profiles generated by machine learning using data collected by the TS from his / her followers.

6. A recruitment content generation device, characterized in that: The generating device comprises: a network interface connecting the generating device to a social network; Memory; and A processor, the processor being coupled to the memory, the processor being configured to: Receive job descriptions from HR clients; processing the job description by identifying a set of target attributes, wherein the set of target attributes includes a formatted list of job requirements, keywords, categories, and associations; Generate job announcements based on job description templates; Generate standardized job descriptions; Generate personalized job descriptions based on talent preferences; Generate personalized job description content based on TS; Push the generated personalized job description content to TS.

7. The generating device according to claim 6, characterized in that The processor is further configured to: before pushing the generated personalized job description content to the TS, receiving input from the HR client; Based on the received input, determining whether the generated personalized job description content needs to be updated; If it is determined that an update is needed, the generated personalized job description content is updated.

8. The generating device according to claim 6 or 7, characterized in that: The processor is further configured to: A TS is selected based on a recruitment request received from the HR client and a set of TS attributes.

9. The generating device according to claim 8, characterized in that The selecting a TS based on the recruitment request received from the HR client and a set of TS attributes comprises: identifying a group of talent scouts TS from a social network, wherein each identified TS is assigned a set of TS attributes, and wherein each TS has a social media influence value above a predefined influence threshold; processing the job request by assigning a set of target attributes based on the job request; A subgroup of TSs is selected from the identified group of TSs by comparing the set of TS attributes of each TS in the group of TSs with the set of target attributes based on one or more predefined matching criteria.

10. The generating device according to claim 9, characterized in that Identifying TS groups involves: executing a crawler program starting from a list of member pages of said social network to collect collected information of TS candidates; processing the collected information using an artificial neural network (ANN) implemented in the computer system to assign the set of TS attributes; as well as Groups of TSs are identified using a combination of predefined selection rules and optional corresponding thresholds, talent behavior profiles generated by machine learning using data collected by the TS from his / her followers.

Images