A real-time job recommendation method and system

Abstract

The present application relates to a kind of real-time position recommendation method and system.The method includes: monitoring the active state of job-seeking user, position and position HR;In response to the active state of job-seeking user meets the recommendation condition, according to the job-seeking user label query and match the position label that meets the active condition to obtain the recommendation information this time, wherein the recommendation information at least includes position information and instant communication invitation link to position HR;And in response to the acceptance operation of instant communication invitation that position HR issues to job-seeking user, establish instant communication connection between the job-seeking user and the position HR, and provide instant communication window.The system includes state monitoring module, first recommendation engine, push module and instant communication module.The present application provides convenient instant communication mode for the job-seeker and recruiter who are online simultaneously, thereby effectively improves the communication efficiency of job-seeker and recruiter, and further improves the recruitment success rate.

Description

Technical Field

[0001] This invention relates to the field of Internet data processing technology, and in particular to a real-time job recommendation method and system. Background Technology

[0002] In the information age, various professional or comprehensive information platforms provide the information needed by both supply and demand users. Taking the job market as an example, professional recruitment platforms simultaneously provide numerous job postings and job search results for both job seekers and employers. The vast majority of job seekers and recruiters choose to find suitable positions through professional recruitment platforms. Typically, both register on recruitment platforms such as job websites and recruitment apps. Job seekers fill out resumes, including personal information, desired positions, and related requirements, while recruiters fill out job postings, including company information, the specific position being advertised, and job requirements. The traditional communication method between job seekers and recruiters is as follows: after finding a matching position on the recruitment platform, the job seeker submits their resume to the resume submission link provided on the platform (usually the email address of the HR (human resources manager) of the position), and then passively waits for contact from the HR, such as by phone or text message. However, this traditional communication model presents several challenges. Job seekers can only obtain information about the position and company from current job postings, which are often limited in content. This makes it difficult for them to determine if a position is a good fit. Furthermore, after submitting their resumes, job seekers cannot predict when or if they will receive a response, or whether their application will be successful. Without feedback, they are forced to spend time searching for other positions. For recruiters, HR only receives resumes when checking their inbox, resulting in poor timeliness. Moreover, after a job posting, HR may receive numerous resumes, but most do not meet their requirements or preferences, requiring time for screening and evaluation. Therefore, this traditional communication method is inefficient and has a low success rate for both job seekers and recruiters. For recruitment platforms, inefficient recruitment services fail to attract more users (job seekers and recruiters) and lead to user churn. Thus, providing efficient recruitment services has always been a goal for recruitment platforms. Summary of the Invention

[0003] To address the technical problems existing in the prior art, this invention proposes a real-time job recommendation method and system to improve communication efficiency between job seekers and recruiters, thereby increasing the success rate of job seeking or recruitment.

[0004] To address the aforementioned technical problems, according to one aspect of the present invention, a real-time job recommendation method is provided, comprising the following steps: monitoring the activity status of job seekers, jobs, and job HRs; in response to a job seeker's activity status meeting recommendation criteria, querying and matching job tags that meet the activity criteria based on the job seeker's tags to obtain the current recommendation information, wherein the recommendation information includes at least job information and an instant communication invitation link to the job HR; and in response to the job HR accepting the instant communication invitation issued by the job seeker, establishing an instant communication connection between the job seeker and the job HR, and providing an instant communication window.

[0005] According to another aspect of the present invention, a real-time job recommendation system is provided, comprising a status monitoring module, a first recommendation engine, a push module, and an instant messaging module. The status monitoring module is configured to monitor the activity status of job seekers, jobs, and job HR personnel, and sends a recommendation trigger notification when the activity status of a job seeker meets recommendation criteria. The first recommendation engine is connected to the status monitoring module and is configured to, upon receiving the recommendation trigger notification, query and match job tags that meet the activity criteria based on job seeker tags to obtain multiple jobs. The push module is connected to the first recommendation engine and is configured to generate recommendation information based on the multiple jobs and push it to the job seeker. The recommendation information includes at least job information and an instant messaging invitation link to the job HR personnel. The instant messaging module is connected to the push module and is configured to, in response to the job HR personnel accepting an instant messaging invitation from the job seeker, establish an instant messaging connection between the job seeker and the job HR personnel and provide an instant messaging window.

[0006] This invention monitors users' online status by observing their behavior, providing a convenient instant messaging mode for job seekers and recruiters who are online simultaneously. This effectively improves communication efficiency between job seekers and recruiters, thereby increasing the recruitment success rate. Attached Figure Description

[0007] The preferred embodiments of the present invention will now be described in further detail with reference to the accompanying drawings, wherein:

[0008] Figure 1 This is a block diagram illustrating the principle of a real-time job recommendation system according to Embodiment 1 of the present invention;

[0009] Figure 2 This is a block diagram of the status monitoring module according to Embodiment 1 of the present invention;

[0010] Figure 3 This is a block diagram of the principle of the first recommendation engine according to Embodiment 1 of the present invention;

[0011] Figure 4 This is a block diagram of another first recommendation engine principle according to Embodiment 1 of the present invention;

[0012] Figure 5 This is a block diagram of the push module principle according to Embodiment 1 of the present invention;

[0013] Figure 6 This is a flowchart of a real-time job recommendation method according to Embodiment 1 of the present invention;

[0014] Figure 7 This is a flowchart of the recommended triggering method according to Embodiment 1 of the present invention;

[0015] Figure 8 This is a flowchart of a method for querying and matching job positions according to Embodiment 1 of the present invention;

[0016] Figure 9 This is a block diagram of the real-time job recommendation system provided in Embodiment 2 of the present invention;

[0017] Figure 10 This is a block diagram of the real-time job recommendation system provided in Embodiment 3 of the present invention;

[0018] Figure 11 This is a partial block diagram of a real-time job recommendation system according to Embodiment 3 of the present invention; and

[0019] Figure 12 This is a block diagram of the principle of a real-time job recommendation system provided in Embodiment 4 of the present invention. Detailed Implementation

[0020] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0021] In the following detailed description, reference can be made to the accompanying drawings, which form part of this application and illustrate specific embodiments of the present application. In the drawings, similar reference numerals describe substantially similar components in different figures. Specific embodiments of the present application are described in sufficient detail below to enable those skilled in the art to implement the technical solutions of the present application. It should be understood that other embodiments may also be utilized, or structural, logical, or electrical changes may be made to the embodiments of the present application.

[0022] With the popularization and development of instant messaging applications, communication applications have greatly improved communication efficiency to some extent, enabling people to communicate about certain events anytime and anywhere. This invention provides a real-time job recommendation method and system. By leveraging instant messaging applications, job seekers and recruiters who are simultaneously online can communicate in real time. This increases the activity of both job seekers and HR personnel, thereby enhancing user stickiness on the platform and increasing the number of job views and applications, ultimately improving the probability of a match between the two parties. The invention will be described in detail below through specific embodiments.

[0023] Example 1

[0024] Figure 1 This is a block diagram illustrating the principle of a real-time job recommendation system according to Embodiment 1 of the present invention. The real-time job recommendation system includes a status monitoring module 1, a first recommendation engine 2, a push module 3, and an instant messaging module 4. In this invention, a behavior collection module 7 records and collects various operational data from job seekers and job HR personnel, as well as the data on the operations performed on each job. For example, by embedding data points at preset locations, the behavior of job seekers and job HR personnel is collected. For job seekers, behaviors such as logging into the platform, searching for jobs, viewing jobs, submitting resumes to a job, and sending instant messaging invitations to a job HR personnel can be collected. For job HR personnel, behaviors such as logging into the platform, searching for or viewing jobs, posting or delisting jobs, sending interview notifications to job seekers who have submitted resumes, and accepting instant messaging invitations can be collected. By monitoring these behaviors of job seekers and job HR personnel, their online status, online time, and duration can be determined. Behavior logs are generated for each operation and behavior of job seekers and job HR personnel and stored in a database 5.

[0025] To facilitate user status queries, the status maintenance module 6 categorizes user statuses based on user behavior and stores the user status information in the database 5. It maintains the user status information in the database whenever new behavioral data is generated. In one embodiment, the status maintenance module 6 divides user statuses into historical statuses and current statuses. Historical statuses include multiple behaviors at specific times, with corresponding statuses such as inactive, active, registered, applied, and registered / applied. For ease of querying, these statuses are represented by codes. "Active" refers to actions such as querying and viewing. Current statuses include details of the day's recommendations, such as recommendation time, frequency, the job postings recommended each time, and the corresponding HR information.

[0026] In addition, each job posting has some basic data, such as the posting date, expiration date, company information, HR information, recent activity time, various job tags, and the current status (posted, under review, or delisted). The status maintenance module 6 modifies some of the above basic data, such as recent activity time and status, based on the HR's actions regarding the job, such as posting, processing resumes submitted, scheduling interviews, and delisting. The job tags include tags for desired positions, work location, industry, job type (full-time, part-time, or internship), required skills, and education level.

[0027] This invention includes a recommendation strategy. The status monitoring module 1 determines whether to send a recommendation trigger notification to the first recommendation engine 2 based on the monitored job seeker status and the corresponding recommendation strategy. In one embodiment, as shown... Figure 2 As shown, the status monitoring module 1 includes a user status determination unit 11, an activity level determination unit 12, and a recommendation triggering unit 13. The user status determination unit 11 is connected to the database 5 or to the user behavior collection module 7. The user status determination unit 11 can determine user status and behavior by periodically checking the user status in the database 5. For example, it queries the user's historical status for a certain period of time; if the user's status is active, submitted, etc., during that period, an activity notification is sent to the recommendation triggering unit 13 and the activity level determination unit 12.

[0028] After receiving the activity notification from the user status determination unit 11, the activity level determination unit 12 queries the user's historical status in the database 5 and determines the user's activity level based on that historical status. Different activity levels receive different recommendation strategies, resulting in variations in the total number of job postings pushed daily, the initial push count, and the recommendation interval. For example:

[0029] Once a user is newly registered, their level is determined to be a Level 1 active user. The daily recommendation limit is 100. The first action triggers a maximum of 10 recommendations, and subsequent actions triggered at intervals of more than 2 minutes will push a maximum of 2 recommendations.

[0030] Users who are inactive for the first 29 days are considered Level 2 active users. The daily recommendation limit is 40. The first behavior triggers a maximum of 5 recommendations. Subsequent recommendations will be pushed if there is a triggering behavior at an interval of more than 2 minutes.

[0031] For users who are active at level 3 within the first 29 days, the daily recommendation limit is 20. The first behavior triggers a maximum of 3 recommendations. Subsequent recommendations will be pushed if there is an interval of more than 5 minutes and there is a triggered behavior.

[0032] After receiving an activity notification from the user status determination unit 11, the recommendation triggering unit 13 obtains the job seeker's activity level through the activity level determination unit 12. It then queries the recommendation details stored in the job seeker's status for the day, selects a corresponding recommendation strategy based on the job seeker's activity level and the current recommendation situation, and sends a recommendation triggering notification to the first recommendation engine 2 when the recommendation strategy is met.

[0033] Upon receiving a recommendation trigger notification, the first recommendation engine 2 queries and matches job tags that match the activity criteria based on the job seeker's tags to obtain multiple job listings. In one embodiment, such as... Figure 3 As shown, the first recommendation engine 2 includes a user tag acquisition unit 21, a query unit 22, an HR online query unit 23, and a sorting unit 24.

[0034] The recommendation trigger notification includes a job seeker's identity identifier, and the user tag acquisition unit 21 retrieves the job seeker's user tag from the database 5 based on the job seeker's identity identifier. The job seeker's user tag includes at least filtering tags for job searching and matching. In one embodiment, the filtering tags include at least a desired job tag and a work location tag; more preferably, they may also include one or more of the following: job title tag, education level tag, years of work experience tag, salary tag, job type tag, and preference tag. The job seeker's user tag also includes scoring tags for sorting, such as one or more of the following: desired job tag, industry tag, skill tag, language tag, course tag, sales channel tag, age tag, medical tag, gender tag, and educational background tag. The sorting tags also include job activity time, allowing multiple jobs to be sorted based on their activity time.

[0035] The HR online query unit 23 queries the HR behavior logs in database 5 to determine the positions that the HR is online for, and to determine whether the corresponding positions are valid, thereby identifying the currently active and valid positions.

[0036] The query unit 22 queries the job tags of currently active jobs according to filter tags, such as desired job tags and work location tags, to obtain a first result job list. However, sometimes there are very few active jobs that match both job intentions and work locations. In one embodiment, to address this situation, the first recommendation engine 2 further includes a tag expansion unit 25. When the query unit 22 retrieves too few jobs, it sends a notification to the tag expansion unit 25, which expands some of the currently used filter tags. For example, when a job seeker has only one desired job tag, that desired job tag can be expanded. In one embodiment, a global statistical analysis is performed based on the job seeker's desired job tag to obtain one or more tags that co-occur with the job seeker's desired job tag and the number of times they co-occur, and the tags that co-occur most frequently with the job seeker's desired job tag are obtained. For example, job level tags can be expanded. For example, the current level can be expanded two levels upwards and one level downwards. In one embodiment, the filter tags that can be expanded include, for example, years of work experience, upper salary limit, and lower salary limit. The tag extension unit 25 sends the extended tags to the query unit 22, which then performs a new query and matching until a sufficient number of jobs are obtained.

[0037] The sorting unit 24 obtains sorting tags from the user tag acquisition unit 21. In one embodiment, the sorting tags include rating tags and job activity time tags. Rating tags include tags for desired jobs, industries, skills, languages, courses, sales channels, ages, medical backgrounds, genders, educational backgrounds, etc. Multiple jobs in the first result job list are scored based on the matching degree between the rating tags in the user sorting tags and the rating tags in the job tags, and then sorted according to the scores to obtain a second result job list.

[0038] like Figure 4 The diagram shown is a partial block diagram of a first recommendation engine 2 according to another embodiment. In this embodiment, the first recommendation engine 2 includes a user tag acquisition unit 21, a query unit 22, an online HR query unit 23, a sorting unit 24, and a tag expansion unit 25. Figure 4In addition to (not shown), it also includes a first filtering unit 26. The query unit 22 performs query matching based on the job seeker filtering tags obtained by the user tag acquisition unit 21, thereby obtaining multiple positions. The HR online query unit 23 queries the database 5 to determine the currently active and valid positions. The first filtering unit 26 compares the multiple positions obtained from the query with the active and valid positions determined by the HR online query unit 23, thereby filtering out inactive positions. The sorting unit 24 scores and sorts the active and valid positions obtained after filtering to obtain a second result position list. In this embodiment, querying and matching are performed first based on user filtering tags, and then inactive positions are removed from the matching results.

[0039] The push module 3 is connected to the status monitoring module 1 to obtain the current activity level of the job seeker and determine the number of positions to be recommended based on the activity level. To avoid repeatedly recommending positions or HR positions to the job seeker, the positions in the obtained position list need to be filtered to remove unsuitable positions. In one embodiment, such as... Figure 5 As shown, the push module 3 includes a job quantity determination unit 31, a second filtering unit 32, and a recommendation information generation unit 33. The job quantity determination unit 31 is connected to the status monitoring module 1 and determines the number of jobs to be recommended based on the user's activity level. For example, if the current job seeker is a newly registered user and this is their first recommendation, then according to the recommendation strategy, the number of jobs recommended this time is 10. If it is not their first recommendation, then the number of jobs recommended this time is 2.

[0040] The second filtering unit 32 has built-in filtering rules, which filter the positions in the second result job list obtained by the first recommendation engine 2 according to the filtering rules. For example, it queries the database 5 to obtain the positions that the job seeker has recommended and applied for within a preset time period (e.g., 14-21 days), and deletes these positions from the second result job list; it queries whether the positions in the second result job list have been rejected by HR, and deletes them if so; it queries whether the positions in the second result job list have the same HR as the recommended positions, and deletes them if so; it queries whether the positions in the second result job list have the same recruiter identity as the recommended positions on the same day, and deletes them if so; it queries whether the positions in the second result job list have the same recruiter identity as the recommended positions within the preset time period and have a historical recommendation count greater than a threshold, and deletes them if so; it queries whether the positions in the second result job list are positions that the user has set to block recruiters from posting, and deletes them if so. After filtering by the second filtering unit 32, a third result job list is obtained.

[0041] The recommendation information generation unit 33 obtains a corresponding number of positions from the third result position list based on the number determined by the position quantity determination unit 31, and obtains the corresponding position information, position HR information, and instant communication invitation link to the position HR to generate recommendation information, which is then pushed to the job seeker.

[0042] Instant Messaging Module 4 is used to establish a communication connection between job seekers and HR personnel, and provides an instant messaging dialog window. The dialog window provides a text input field, an emoticon field, a file transfer field, and a voice / video connection button, facilitating text, voice, and video communication between job seekers and HR personnel, and also meeting file transfer needs.

[0043] Figure 6 This is a flowchart of a real-time job recommendation method according to Embodiment 1 of the present invention. (In conjunction with the foregoing...) Figure 1-5 The principle block diagram is shown below, and the method is explained as follows:

[0044] Step S1: Monitor the activity status of job seekers, job postings, and HR personnel. Based on the job seeker's status and the corresponding recommendation strategy, send a recommendation trigger notification to the first recommendation engine 2. In one embodiment, the recommendation trigger process is as follows: Figure 7 As shown:

[0045] In step S11, the status monitoring module 1 monitors the user's status. The user status determination unit 11 in the status monitoring module 1 can obtain the current active status of the job seeker by viewing the user status in the database 5; or, the user status determination unit 11 is connected to the user behavior collection module 7 and receives notifications sent by it. For example, when the user behavior collection module 7 collects specified user behaviors (such as login, viewing, and search behaviors), it sends a notification to the user status determination unit 11 while writing behavior logs to the database.

[0046] Step S12: Determine if a specified behavior has occurred. For example, whether a notification has been received from the user behavior collection module 7, or whether a change in the user's status for the day has been found in the database 5. If no specified behavior such as job search, query, or application has occurred for the user, return to step S11 to continue monitoring. If a specified behavior has occurred, proceed to step S13.

[0047] Step S13: Obtain the current job seeker's activity level. Different activity levels result in different total number of job postings pushed daily, different initial push counts, and different recommendation intervals. The user's activity level can be determined by querying their historical status. In one embodiment, user activity levels are divided into three levels: for example, newly registered users are Level 1 active users, users who were inactive for the first 29 days are Level 2 active users, and users who were active for the first 29 days are Level 3 active users.

[0048] Step S14: Determine whether a job recommendation has been made to the user that day. If not, this indicates that the action occurred for the first time that day, and a recommendation trigger notification is sent in step S17. If a job recommendation has already been made to the user that day, proceed to step S15.

[0049] Step S15: Query the last job posting time in the recommended details stored in the current day's status, then calculate the interval between the last post and the current time. Determine if this interval is greater than a preset time interval, such as 2 minutes. If it is, proceed to step S16. If the interval between the last post and the current time is less than the preset time interval, to avoid disturbing job seekers with excessively frequent pushes, no push is made, and the process returns to step S11 to monitor the next user behavior. Alternatively, wait until the interval between the last post and the current time is greater than the preset time interval before proceeding to step S16.

[0050] Step S16: Determine if the maximum number of recommended positions has been reached. If the maximum number of recommendations has been reached, end the recommendation process. Otherwise, send a recommendation trigger notification in step S17.

[0051] Step S2: Obtain the job seeker tags. The job seeker tags include at least filtering tags for job searching and matching, and scoring tags for sorting.

[0052] Step S3: Based on the job seeker's filter tags, query and match job postings to obtain a first result job list including multiple positions. In one embodiment, the specific query and matching process is as follows: Figure 8 As shown.

[0053] Step S31: Query the HR behavior logs and job status data in database 5 to identify active jobs that are valid and whose HR personnel are online.

[0054] Step S32: Query the job tags of currently active jobs according to the filter tags, such as the desired job tag and the work location tag, to obtain the first result job list.

[0055] Step S33: Determine whether the number of positions in the first result job list is greater than or equal to the number of positions recommended this time. If not, proceed to step S34. If the number of positions recommended this time is reached, proceed to step S35.

[0056] Step S34: Expand some of the filter tags in the job seeker tags, and then return to step S32.

[0057] Step S35: Sort the jobs in the first result job list. In one embodiment, the jobs are scored according to the matching degree between job seekers and jobs based on the scoring tags, such as desired job tags, industry tags, skill tags, language tags, course tags, sales channel tags, age tags, medical tags, gender tags, education background tags, etc., and then sorted by score to obtain a second result job list. In another embodiment, multiple jobs are sorted according to their activity time.

[0058] Step S4 involves filtering the query and matching results. For example, filtering out jobs that have been recommended or applied for within a preset time period (e.g., 14-21 days) from the second result job list; filtering out jobs rejected by HR; filtering out jobs with the same HR as recommended jobs; filtering out jobs with the same recruiter identity as recommended jobs on the same day; filtering out jobs with the same recruiter identity as recommended jobs within the preset time period and with a historical recommendation count exceeding a threshold; filtering out jobs posted by recruiters blocked by users, etc. After the above filtering, the third result job list is obtained.

[0059] Step S5: Generate recommendation information and push it to the job seeker. Specifically, based on the number n jobs recommended this time, extract the top n jobs from the third result job list, generate recommendation information by including the job information, HR information, and instant communication invitation link for the HR for each of the n jobs, and push this information to the job seeker.

[0060] Step S6: Monitor the behavior of the job seeker and the HR personnel in the recommended information. This includes, for example, monitoring the job seeker's actions on the recommended information and job postings, such as viewing job information, sending an instant messaging invitation link to the HR personnel, and closing the recommended information; simultaneously monitoring the HR personnel's responses to the instant messaging invitation.

[0061] Step S7: Determine whether the job seeker has turned off the recommendation information. If the recommendation information has been turned off, end the current recommendation process and return to step S1. Otherwise, proceed to step S8.

[0062] Step S8: Determine whether the job seeker has sent an instant communication invitation. If the job seeker has sent an instant communication invitation, in step S9, send the job seeker's personal tag information to the invited HR. If the job seeker has not sent an instant communication invitation, return to step S6.

[0063] Step S10: Determine whether the invited HR has accepted the invitation. If the invited HR has accepted the invitation, in step S11, establish an instant messaging connection between the job seeker and the HR and provide an instant messaging window. If the invited HR has declined the invitation, record the action and return to step S6.

[0064] Step S12: Determine whether the instant messaging connection between the job seeker and the HR of the position is broken. If it is broken, end the instant messaging and return to step S6. If it is not broken, proceed to step S12.

[0065] In this embodiment, when job seekers are online, positions with online HR personnel are recommended to them, enabling them to communicate with HR personnel online via instant messaging, thus improving communication efficiency between job seekers and recruiters. The recommended strategy ensures that positions are recommended to job seekers at an appropriate frequency without causing interference. When searching for positions for job seekers, tags representing user intent from various angles are used for searching and matching, resulting in a higher degree of job matching that more closely reflects the user's true intent.

[0066] Example 2

[0067] Figure 9 This is a block diagram illustrating the principle of a real-time job recommendation system according to Embodiment 2 of the present invention. The difference between this embodiment and Embodiment 1 is that this embodiment further includes a preference update module 8. In this embodiment, the preferences of job seekers and job HRs are obtained based on monitored job seeker behavior and HR behavior. After pushing recommendation information to job seekers, the behavior collection module 7 collects the operation information of job seekers and job HRs on the recommended positions and stores it in the database 5. The present invention determines the commonalities of these positions based on the positions viewed and applied for by job seekers, such as the same region, the same level, the intersection of salary ranges, etc., thereby determining the various preferences of job seekers in terms of region, job level, salary, industry, etc. When recommendation information is sent, the preference update module 8 monitors the job seeker's operation behavior on the recommended positions, such as viewing or not viewing, sending instant messaging invitations to job HRs, saving the position, applying for the position, etc., and updates the job seeker's preference tags based on this operation information and historical behavior data. Similarly, the preference update module 8 recalculates the HR's preferences for the position based on the positions viewed by the HR, the interview notices sent, and the instant messaging invitations rejected, such as preferences for education, years of work experience, and skills.

[0068] When performing queries and matching, First Recommendation Engine 2 not only uses filter tags for query matching, but also uses job seeker preferences and HR preferences for job openings to determine positions that match both the job seeker's preferences and the HR's preferences, thereby increasing the recommendation success rate.

[0069] In another embodiment, the ranking of the query results from the first recommendation engine 2 is determined using the preferences of job seekers and the preferences of job HR personnel. In this embodiment, the first recommendation engine 2 still performs query matching based on the filter tags of job seekers. After obtaining the query results, multiple positions in the query results need to be ranked. In this embodiment, the weight of each scoring tag is determined based on the user's preferences. For example, the industry tags for job seekers include three categories: "Computer Software," "Computer Hardware," and "Internet." Based on the user's behavior in job searching, viewing, and applying, it is determined that the user prefers "Internet," followed by "Computer Software," and lastly "Computer Hardware." Based on statistical analysis of this behavior, the user's weight values ​​for these three industries are determined as follows: "Computer Software" 0.3, "Computer Hardware" 0.2, and "Internet" 0.5. When the first recommendation engine 2 retrieves positions that include these industries, the industry tag score is multiplied by its respective weight, thus reflecting the user's preference in that industry category. Similarly, the preferences of job HR personnel are also reflected in the corresponding tag weights during scoring. For example, the weight value is determined based on the difference between the user's and HR's preferences. For example, if an HR professional prefers candidates with a master's degree or higher for a particular position, and the user's education level meets the requirements (master's degree or higher), then the weight of the education tag is set to 1. If the user's education level is a bachelor's degree and does not meet the requirements for a master's degree or higher, then the weight of the education tag is set to 0.8. This way, positions that better match the HR professional's preferences will be ranked higher during the sorting process.

[0070] Example 3

[0071] Figure 10 This is a block diagram illustrating the principle of a real-time job recommendation system according to Embodiment 3 of the present invention. In this embodiment, the real-time job recommendation system further includes a behavior monitoring module 9 and a second recommendation engine 10. The behavior monitoring module 9 is connected to the database 5 and monitors a preset number of user submission and / or viewing behaviors. For example, if a user submits 5 times and views 5 times during their current online session, it meets the recommendation criteria, and the jobs involved in the aforementioned 10 behaviors are sent to the second recommendation engine 10. The second recommendation engine 10 queries similar jobs based on these jobs and filters the query results based on the activity status of the user, job, and job HR obtained from the status monitoring module 1 to remove jobs where the HR is offline or invalid. Specifically, for example... Figure 11As shown, the behavior monitoring module 9 includes a behavior monitoring unit 91 and a job acquisition unit 92. The behavior monitoring unit 91 checks the application and viewing behavior logs of job seekers in the database in real time or periodically. If a job seeker's recent application and / or viewing behavior meets the requirements, such as 5 recent applications and / or 5 recent views, a job acquisition notification is sent to the job acquisition unit 92. After receiving the job acquisition notification, the job acquisition unit 92 queries the database 5 according to the job seeker's identity identifier in the notification, retrieves the jobs involved in the job seeker's application and / or viewing behavior, generates a job list, and sends it to the second recommendation engine 10. The second recommendation engine 10 includes a similar job calculation unit 101, a third filtering unit 102, and an HR online query unit 103. The job acquisition unit 92 sends the generated job list to the similar job calculation unit 101. The similar job calculation unit 101 queries the job database, uses relevant algorithms to obtain one or more jobs similar to a job in the list, generates a similar job list, and sends it to the third filtering unit 102. The third filtering unit 102 is connected to the status monitoring module 1 to obtain the activity level of the job seeker. Simultaneously, the HR online query unit 103 queries the database for the behavior data of job HR personnel, identifies positions where the HR personnel are online and valid, and sends a list of such positions to the third filtering unit 102. The third filtering unit 102 filters the positions in the similar position list based on validity and activity, removing invalid positions and positions where the HR personnel are offline. Based on the user level and recommendation strategy, it determines a corresponding number of positions to obtain the final similar position list, which is then sent to the push module 3. The push module 3 generates recommendation information based on the similar position list and pushes it to the user. The process of establishing instant communication with the job HR personnel is similar to that in Embodiment 1 and will not be described again here.

[0072] This embodiment provides users with recommended jobs from another perspective, expanding the breadth of job recommendations and thus meeting the needs of job seekers in multiple ways, effectively improving the job search success rate.

[0073] Example 4

[0074] The system principle block diagram of this embodiment is as follows: Figure 12 As shown, this embodiment includes both the first recommendation engine 2 of Embodiment 1 and the second recommendation engine 10 of Embodiment 3. When the status monitoring module 1 detects that a user is active, it selects a corresponding recommendation strategy based on the job seeker's activity level and the current recommendation status. When the recommendation strategy is met, a recommendation trigger notification is sent to the first recommendation engine 2 and the second recommendation engine 10 respectively. The first recommendation engine 2 obtains the second result job list according to Embodiment 1 and sends it to the push module 3, which will not be described in detail here.

[0075] When the behavior monitoring module 9 detects that the user's behavior meets the preset conditions (the user has recently made 5 submissions and 5 views), it triggers the second recommendation engine 10 to query similar positions, obtain a list of similar positions, and store it in the database. When the second recommendation engine 10 receives the recommendation trigger notification sent by the status monitoring module 1, it retrieves the most recently obtained list of similar positions from the database and sends it to the push module 3.

[0076] The user push module 3 merges the second result job list obtained by the first recommendation engine 2 and the similar job list obtained by the second recommendation engine 10, then filters the merged job list, and then extracts the top-ranked jobs from the filtered job list according to the number of jobs to be recommended this time to generate recommendation information.

[0077] The process of establishing instant messaging with the HR of the position is similar to that in Example 1, and will not be repeated here.

[0078] This invention monitors users' online status by observing their behavior, and combines this with different recommendation strategies to recommend relevant positions to online job seekers who are also online with HR personnel. It also provides a convenient instant messaging mode, thereby effectively improving the communication efficiency between job seekers and recruiters, and increasing the job search success rate and recruitment success rate.

[0079] The above embodiments are for illustrative purposes only and are not intended to limit the invention. Those skilled in the art can make various changes and modifications without departing from the scope of the invention. Therefore, all equivalent technical solutions should also fall within the scope of the invention.

Claims

1. A real-time job recommendation method, comprising: User status is categorized based on user behavior and stored in a database. The user status information in the database is maintained when new behavioral data is generated. User status is divided into historical status and current status. Historical status includes multiple behaviors at specific times, and the status corresponding to the behaviors is divided into inactive, active, registered, submitted, and registered and submitted. Monitor the activity status of job seekers, job postings, and job HR personnel. Specifically, the activity status of job seekers is obtained by viewing the user status in the database and their activity level is determined. Different activity levels result in different total number of job postings pushed per day, number of initial pushes, and recommendation intervals. In response to the job seeker's activity status meeting the recommendation criteria, a set of job tags that meet the activity criteria is obtained. Based on the job seeker's filter tags, job queries and matching are performed in the set of job tags to obtain a first result job list. The activity criteria are that the job is valid and its HR is online. The recommendation criteria include: detecting the job seeker's first action on the day; or, detecting that the job seeker's non-first action on the job on the day, and the number of recommendations to the job seeker on the day is less than the upper limit threshold, and the time interval between the last recommendation and the last recommendation reaches the threshold. The scores of multiple jobs in the first result job list are scored based on the matching degree between the score tags in the user sorting tag item and the score tags in the job tag item. The matched jobs are then sorted according to the score results and the job activity time to obtain the sorted second result job list. The query and matching results are filtered out from the second result job list, removing jobs that have been recommended and applied for within a preset time period, jobs that have been rejected by HR, jobs with the same HR as the recommended job, jobs with the same recruiter identity as the recommended job on the same day, jobs with the same recruiter identity as the recommended job within the preset time period and with a historical recommendation count greater than a threshold, and jobs posted by recruiters that have been blocked by users, to obtain the third result job list. Extract the top n jobs from the third result job list, generate recommendation information by combining the job information, HR information, and instant communication invitation links for the HR for these n jobs, and push this information to the job seeker; and Monitor the job seeker's actions on the recommended information, and the HR's response to the instant communication invitation; Determine whether the job seeker has sent an instant communication invitation. In response to the job seeker's action of sending an instant communication invitation to the HR of the position, send the job seeker's personal tag information to the invited HR. Determine whether the invited HR has accepted the invitation. In response to the HR's acceptance of the instant messaging invitation from the job seeker, establish an instant messaging connection between the job seeker and the HR, and provide an instant messaging window.

2. The real-time job recommendation method according to claim 1, wherein when querying and matching jobs based on user filter tags, if the number of jobs obtained is less than the number of jobs recommended this time, some or all of the filter tag content in the user filter tag items are expanded.

3. The real-time job recommendation method according to claim 1, wherein the filtering tags include one or more of the following tags: desired job tag, work location tag, job level tag, education tag, years of work experience tag, salary tag, job type tag, and preference tag; The scoring tags include one or more of the following tags: desired job tag, industry tag, skills tag, language tag, course tag, sales channel tag, age tag, medical tag, gender tag, and educational background tag; The sorting tags also include job activity time; multiple jobs are sorted based on job activity time.

4. The real-time job recommendation method according to claim 1, further comprising: In response to the job seeker's active status meeting the recommendation criteria, similar positions are calculated based on the positions involved in the job seeker's behavior. This recommendation information is generated based on the similar positions and / or positions matched according to the job seeker's tags.

5. The real-time job recommendation method of claim 1, wherein further comprising: Update the job seeker's preference tags based on the job seeker's interaction information and historical behavior data with the recommended positions; The HR's preference tags are updated based on the job seeker's personal tag information, response information to instant messaging invitations, and historical behavior data.

6. A real-time job recommendation system, comprising: The status maintenance module is configured to classify user status according to user behavior and store user status information in the database. When new behavior data is generated, the module maintains the user status information in the database. The user status is divided into historical status and current status. The historical status includes multiple behaviors at specific times, and the status corresponding to the behaviors is divided into inactive, active, registered, submitted, and registered and submitted. The status monitoring module is configured to monitor the activity status of job positions and their HR personnel, and obtain the activity status of job seekers by checking the user status in the database. When the activity status of a job seeker meets the recommendation criteria, a recommendation trigger notification is sent. The recommendation criteria include: detecting the job seeker's first action on the day; or, detecting a job seeker's non-first action on the job on the day, and the number of recommendations to the job seeker on the day is less than the upper limit threshold, and the time interval between the last recommendation and the previous recommendation reaches the threshold. The first recommendation engine, which is connected to the status monitoring module, is configured to query and match job tags that meet the activity criteria based on job seeker tags to obtain multiple jobs when a recommendation trigger notification is received. A push module, connected to the first recommendation engine, is configured to generate recommendation information based on the multiple job positions and push it to the job seeker. It determines whether the job seeker has issued an instant communication invitation. In response to the job seeker's action of issuing an instant communication invitation to the job HR, it sends the job seeker's personal tag information to the invited HR. The recommendation information includes at least job information and an instant communication invitation link to the job HR. An instant messaging module, connected to the push module, is configured to respond to an instant messaging invitation sent by a job seeker when the HR of the position accepts the invitation, establish an instant messaging connection between the job seeker and the HR, and provide an instant messaging window; The status monitoring module includes: The user status determination unit is configured to monitor the current activity status of job seekers and send an activity notification when the job seeker's current status is active. An activity level determination unit, which is connected to the user status determination unit, determines the activity level of the job seeker upon receiving an activity notification for the job seeker. The first recommendation engine includes: The user tag acquisition unit is configured to acquire job seeker tags based on the job seeker identity identifier in the recommendation trigger notification; wherein, the job seeker tags include filter tags for job search and matching and tags for sorting; The HR online query unit, once configured, identifies currently active and valid HR positions by querying HR behavior data. A query unit, connected to the user tag acquisition unit and the HR online query unit, is configured to perform job searches and matching within currently active and valid HR positions based on the job seeker's filter tags, to obtain a first result job list; and A sorting unit, which is connected to the user tag acquisition unit and the query unit, is configured to sort multiple job positions obtained from the query according to the sorting tag items to obtain a second result job list; The push module includes: A job posting quantity determination unit, which is connected to the activity level determination unit, is configured to determine the number of job postings recommended this time based on the activity level of the job seeker; The filtering unit is configured to remove from the current second-result job list jobs that have been recommended and applied for within a preset time period, jobs rejected by HR, jobs with the same HR as recommended jobs, jobs with the same recruiter identity as recommended jobs on the same day, jobs with the same recruiter identity as recommended jobs within the preset time period and with a historical recommendation count exceeding a threshold, and jobs posted by recruiters blocked by the user, in order to obtain the third-result job list; and The recommendation information generation unit is connected to the job quantity determination unit and the filtering unit to obtain a corresponding number of jobs from the third result job list based on the number of jobs recommended this time, and to obtain the corresponding job information, job HR information, and instant communication invitation link to the job HR to generate recommendation information and push it to the job seeker.

7. The real-time job recommendation system according to claim 6, wherein the status monitoring module further comprises: The recommendation triggering unit is connected to the user status determination unit and the activity level determination unit respectively. When it receives an activity notification for a job seeker, it queries the user's activity status and activity level, and sends a recommendation triggering notification when the job seeker's activity status meets the recommendation conditions.

8. The real-time job recommendation system according to claim 6, wherein the first recommendation engine further includes a tag expansion unit connected to the query unit, configured to expand some or all of the tag content in the filter tag items of job seekers when the number of jobs obtained by the query unit is less than the number of jobs recommended this time.

9. The real-time job recommendation system according to claim 6, further comprising a preference update module, configured to update the job seeker's preference tags based on the job seeker's operation information and historical behavior data in the recommended information; update the job HR's preference tags based on the job seeker's operation information on the job seeker's personal tag information, response information to instant messaging invitations, and historical behavior data; and when the first recommendation engine queries and matches job tags that meet the activity criteria, it performs matching between user preferences and job HR preferences.

10. The real-time job recommendation system according to claim 6, further comprising a second recommendation engine connected to the status monitoring module and the push module, wherein upon receiving a recommendation trigger notification, the engine performs matching calculations based on the jobs involved in the job seeker's behavior to obtain a list of similar jobs, and sends the list of similar jobs to the push module.

Images