WO2016076790A1

WO2016076790A1 - Method and system for profiling job candidates

Info

Publication number: WO2016076790A1
Application number: PCT/SG2015/000023
Authority: WO
Inventors: Annie Hui Lian YAP; Jacob Chiap Ling LIEU; Cheng Chun PHUA
Original assignee: Life Science Board Pte. Ltd.
Priority date: 2014-11-14
Filing date: 2015-01-28
Publication date: 2016-05-19
Also published as: TW201617981A

Abstract

The present invention relates to recruitment systems, in particular, a method and system method for profiling job candidates using searchable information contained in a candidate's resume, the method including determining a list of keywords associated with a job role using a processor, allocating a keyword relationship rating to a candidate based upon a comparison of information in the candidate's resume to the keywords associated with the job role and using a processor, determining candidate classification data by classifying each candidate into one of a plurality of predefined categories based upon the candidate's keyword relationship rating, the categories defining the likelihood of the candidate being suitable for the job role; and storing the candidate classification data in a database.

Description

METHOD AND SYSTEM FOR PROFILING JOB CANDIDATES

FIELD OF THE INVENTION

The present invention relates to recruitment systems, in particular, a method and system for profiling job candidates for use in determining suitable job candidates for a job role.

BACKGROUND TO THE INVENTION

There are existing job boards/portals on the Internet that facilitate employers to post job advertisements to attract potential candidates to apply for the jobs. The job boards/portals usually allow candidates to search for the potential job opportunities using job titles, company names, skillsets and/or keywords to filter out the interested job opportunities for each candidate based on his/her input.

These portals also allow employers to search for potential candidates and review potential candidate's resumes through the employer's databases of registered - candidates/members. The search process requires the employers to provide keywords and their Boolean relationship between the keywords to allow the portal's search engine to provide a filtered list of potential candidates. Employers then need to go through each and every one of the filtered candidates' resumes to decide who to shortlist for interview.

Using such systems means that the quality and success of each filtered list of potential candidates depends on (i) the quality of the candidates in the database, (ii) the completeness and accuracy of the resumes of each candidate in the database, and (iii) the correct keywords and their Boolean relationship used as the search/filter criteria. The third factor is very dependent on the capability and selection of the employer's input into the system, which is often very time consuming and can be rather tricky.

It is therefore desirable to provide an improved method and system for determining suitable job candidates that reduces the time and technical input needed by recruiters.

SUMMARY OF THE INVENTION

In one aspect, the present invention provides a method for profiling job candidates using searchable information contained in a candidate's resume, the method including: determining a list of keywords associated with a job role; using a processor, allocating a keyword relationship rating to a candidate based upon a comparison of information in the candidate's resume to the keywords associated with the job role; and using a processor, determining candidate classification data by classifying each candidate into one of a plurality of predefined categories based upon the candidate's keyword relationship rating, the categories defining the likelihood of the candidate being suitable for the job role; and storing the candidate classification data in a database.

In a second aspect, the present invention provides a method for determining potential job candidates for a job role using a processor including: providing an interface for selecting a job role from a predetermined list of job roles; and displaying candidate classification data stored in a database, said data comprising details of one or more candidates that have been profiled as potentially suitable for the selected job role, wherein the candidate has been profiled by allocating a keyword relationship rating to a candidate based upon a comparison of information in the candidate's resume to predetermined keywords associated with the job role, and classifying each candidate into one of a plurality of predefined categories based upon the candidate's keyword relationship rating, the categories defining the likelihood of the candidate being suitable for the job role. ^*

Allocating the keyword relationship rating may include categorising the keywords into one or more rules, each rule requiring the presence of one or more keywords in the candidate's resume, and based upon said presence of the one or more keywords, allocating a percentage keyword relationship rating for each rule.

The total percentage keyword relationship rating for all rules for a particular job role may be 100%.

Preferably, the method includes providing the candidate classification data to a recruiter.

In a third aspect, the present invention provides a system for profiling job candidates using searchable information contained in a candidate's resume including: determining means for determining a list of keywords associated with a job role; a processor for allocating a keyword relationship rating to a candidate based upon a comparison of . information in the candidate's resume to the keywords associated with the job role, and for determining candidate classification data by classifying each candidate into one of a plurality of predefined categories based upon the candidate's keyword relationship rating, the categories defining the likelihood of the candidate being suitable for the job role; and a database for storing the candidate classification data.

In a fourth aspect, the present invention provides a system for determining potential job candidates for a job role including: an interface for selecting a job role from a predetermined list of job roles; a display for displaying candidate classification data stored in a database, said data comprising details of one or more candidates that have been profiled as potentially suitable for the selected job role, wherein the candidate has been profiled by a processor allocating a keyword relationship rating to a candidate based upon a comparison of information in the candidate's resume to keywords associated with the job role, and determining candidate classification data by classifying each candidate into one of a plurality of predefined categories based upon the keyword relationship rating, the categories defining the likelihood of the candidate being suitable for the job role.

The system may further include the processor categorising the keywords into one or more rules, each rule requiring the presence of one or more keywords in the candidate's resume, and based upon said presence of the one or more keywords, allocating a percentage keyword relationship rating for each rule.

Preferably, the system further includes an output display for providing the candidate classification data and the resumes associated with the candidate classification data to a recruiter.

In another aspect, there is provided a system for determining . potential job candidates for a job role including: an interface for selecting a job role from a predetermined list of job roles; a display for displaying similar candidate data stored in a database, said similar candidate data comprising details of one or more candidates that have been profiled by a processor as a candidate similar to one or more referenced candidates, said referenced candidates comprising candidates who have been profiled as potentially suitable for the job role using the system defined above, wherein said similar candidate data is determined by the similar candidate matching at least two job roles of said reference candidate.

Forms of the present invention provide an improved method and system for determining suitable job candidates that reduces the time and technical input needed by recruiters because the job candidates are already profiled in the system. All that is required by a recruiter is to select a job role from a predetermined list of job roles, and the most suitable candidates are then displayed for review. Complicated Boolean operators do not need to be inputted or selected by recruiters, nor is it necessary for a time consuming search of a lengthy resume database to be undertaken for every search event. The system provides a more reliable and accurate means for profiling candidates as it does not rely on human preferences, e.g. the type of font used in a person's resume. Rather, the system relies solely on predetermined keywords being present in a resume, those keywords having been preselected such that a recruiter need not spend the time selected or searching such keywords him/herself. The output of the system is relatively static - that is, it may be pre-determined for specific job roles such that at the time the recruiter requires candidates, the information is readily available, saving on processing time and reliability of communication and hardware systems.

BRIEF DESCRIPTION OF THE DRAWINGS

An illustrative embodiment of the present invention will now be described with reference to the accompanying figures. Further features and advantages of the invention will also become apparent from the accompanying description.

Figure 1 is a schematic overview of the profiling system of one embodiment of the invention;

Figure 2 is a system diagram showing the hardware components of one embodiment of the invention; and

Figure 3 illustrates the key components of the profiling system embodiment. DESCRIPTION OF PREFERRED EMBODIMENT

The following description is presented to enable any. person skilled in the art to make and use the invention, and is provided in the context of a particular application and its requirements. Various modifications to the disclosed embodiment will be readily apparent to those skilled in the art, and the general principles defined herein may be applied to other embodiments and applications without departing from the spirit and scope of the present invention. Thus, the present invention is not intended to be limited to the embodiments shown, but is to be accorded the widest scope consistent with the principles and features disclosed herein.

Figure 1 is a schematic overview of the profiling system in accordance with one embodiment. At step 1 , the system operator, which may be a recruiter, employer, or independent job candidate agency (hereinafter referred to as 'recruiter' but may encompass any such operator), chooses a job role from a predetermined dropdown menu in a computer program. The program is preferably web-based and run on a web-server (processor), but may be installed as a program on a standalone processor or server. At step 2, the system accesses a database to obtain a list of suitable candidates for the chosen job role which have been profiled using the classification algorithm further described below. At step 3, the list of matched candidates and their resumes, further described below, is provided. The list will be sorted in an order that lists the best-fit candidates at the top of the list. A list of similar candidates can also be listed if a shortlisted/reference candidate is chosen and the recruiter click on the 'Search for Similar Candidates' button.

Figure 2 is a system diagram showing the hardware components of one embodiment of the invention. A candidate uploads a resume 4 to the system via communication means 5 such as the Internet, using a PC, laptop, smart phone or mobile device via web browser(s) or mobile apps or email. Preferably, the resume is uploaded in a format suitable for direct searching of text, for example Adobe or Microsoft Word. The system stores the resume in memory in a database 6. The profiling process is accomplished by the software installed on a processor 8 that may be running all the time in the background. In the profiling process, the resume is searched for certain keywords, described further below. If the resume is not in a directly searchable format then an OCR program or scanner may be used to convert the resume into such searchable format. Once the resume has been searched for keywords, the processor 8 undertakes the steps in the profiling method to classify and rate the candidate associated with the resume, described further below. Once the profiling has completed, the candidate rating, comprising the JC table (described below), is stored in a database, which may be the same database that stores the resume 6, or a standalone database (not shown in figure).

When a recruiter desires to find suitable candidates for a job role, recruiter accesses a computer program stored on a server 8, preferably via the Internet 5 using a PC, laptop, smart phone or mobile device via web browser(s) or mobile apps, and using a keyboard, mouse, touch screen or other input device 7, .selects a job role, described -further below. The processor 8 retrieves, the relevant JC table from the database and displays JC table (or the data stored in the JC table displayed in another desired format) and resumes to the recruiter via a display screen 7. The JC table and resumes may be viewed directly on the display screen 7 via the Internet based computer program, or may be emailed or otherwise provided to the recruiter via communication means 5. -

When a recruiter desires to find similar candidates with reference to a shortlisted candidate, the recruiter accesses a computer program stored on a server 8, preferably via the Internet 5 using a PC, laptop, smart phone or mobile device via web browser(s) or mobile apps_, and using a keyboard, mouse, touch screen or other input device 7, choose a candidate and then click on the 'Search for Similar Candidates' button to request for candidates and their resume with similar profiles. The processor 8 searches the relevant JC table from the database for candidates that hold similar job profiles as the shortlisted candidate and displays the result and resumes to the recruiter via a display screen 7. The similar candidate result and resumes may be viewed directly on the display screen 7 via the Internet based computer program, or may be emailed or otherwise provided to the recruiter via communication means 5.

Figure 3 illustrates the key components of the profiling system embodiment, comprising the Keywords Relationship Weightage (KRW) Table, the Job-Weightage Classification (JWC) Table, and the Job Candidate Classification (JC) Table. These components define how the candidates are classified and rated in the database. Component 1 : Keywords Relationship Weightage (KRW) Table

The first component allocates a keyword relationship rating to a candidate based upon a comparison of information in the candidate's resume to the keywords associated with the job role.

The KRW table is where the rules are defined (unlimited) to how each candidate will be profiled/fit to each job role/function. The basis of the KRW is based on three parts being the keywords, their relationship and their weightage of importance. ^{* '}

As a first step, the system will go through each and every candidate resume at the time of receipt or commencement of the system using the KRW table searching for predefined keywords. The keywords may be the use of word(s), a sentence(s), symbol(s), number(s) and abbreviation(s).

The KRW table 20 in figure 3 shows an example. For Job 1 (21 ), the system has predefined keywords Word 1 , Word2, Word3, Word4, WordS or Word6 (22) associated with that job. A candidate's resume will be scanned to determine if the text of the resume includes one or more of those predefined keywords. For example, is the job is for a 'nurse', then the predefined keywords in Rule 1 may include Nurse, Paramedic, Nursing Manager, Medical, and/or Nursing Director.

The relationship between those keywords is defined in the KRW table 20 in the fourth column 23. In Rule 1 , the relationship between the keywords is OR, meaning that the eventual weight percentage 24 will be based upon whether the resume includes one or more of the keywords. Rule 4 in the KRW table 20 defines the relationship as AND, meaning that the eventual weight percentage 24 will be based upon whether the resume includes all of the keywords defined in Rule 4.

If a candidate meets Rule 1 , the system will go to Rule 2, 3, 4 and so on until all the rules have been considered (in the KRW table 20, being Rule 8). Take Rule 4 for example, the system will go through his resume to search for Word19 and Word 20 and Word21. If all three keywords are found in his resume, the candidate will be assigned another 10% to his Job 1 KRW%. If any of the Word19, Word20 or Word21 is not present in his resume, he will be assigned 0% for Job 1 Rule 4.

Each rule may have a different focus and so define a different aspect of the desired candidate. For example. Rule 1 may be keywords associated with the candidate's job title (e.g. nurse, paramedic, medical, etc). Rule 2 may be keywords associated with the educational levels or qualifications of the candidate (e.g. Bachelor degree, Masters Degree, PhD, Juris Doctor, etc). Rule 3 may be keywords associated with work skills of the candidate (egg good communication skills, word processing skills, Outlook, etc). The number and type of rules are unlimited and will depend upon the particular job role.

The weight percentage associated with each rule is dependent upon the job role. In some embodiments, the weight percentage may be tailored for the needs of a particular recruiter, however it is preferred that the information and results stored in the system is static for each recruiter, minimising the complexities and time involved in the Candidate profiling.

Whether or not a candidate is allocated the weight percentage associated with a particular rule is dependent upon whether the keyword(s) are found in the candidate's resume. If the OR rule is met by one of the keywords in a rule being found, then the full percentage score will be allocated for that rule. If the OR rule is not met as none of the keywords in a rule is found, then a score of 0% will be allocated for that rule. If the AND rule is met by the resume including all of the defined keywords, then the full percentage score will be allocated for that rule. If the AND rule is not met as not all of the keywords in the rule are round, then a score of 0% will be allocated for that rule. In some embodiments, a score between 0% and the maximum percentage score may be allocated, for example in an AND rule where only some of the keywords are located, however this is not preferred for simplification of the system.

Each of the rules is given a predetermined rule weightage depending upon the importance of the particular terms for a particular job. For example, if the job is for a 'nurse' then having the terms Nurse, Paramedic, Nursing Manager, Medical, and/or Nursing Director may be very important and given a 20% rating where one or more of those terms are included, whereas further keywords, for example 'word processing', 'computer' that may appear in Rule 8 for example may be less important and given a 10% rating if one or more of those terms are included. The total weighting for the job rules for a particular job adds up to 100% (25).

Once the resumes have been scanned, and the candidates classified and a weighting given for each of the rules applying to a particular job, the candidate is given a total KRW % for each job. If a rule is deemed compulsory (defined in sixth column 26), the rule must be met or else the candidate will be given 0% for each row and for the total KRW% for that particular job. If a candidate matches no keywords/relationship then they will be given a 0% score.

The process may then be repeated for Job 2 with Job 2 KRW% starting as

0%. By going through each rule, each keyword is being searched for and matched and be given individual rule KRW% value if the keyword-relationship is met and be accumulated towards each job's total KRW%.

The candidate will then be rated with a Job 2 KRW% and so on until all the Jobs and Rules are completely tested.

The candidate's KRW% score for each job is stored in a database.

Component 2: Job-Weightage Classification (JWC) Table

. The JWC table 30 is where the system determines candidate classification data by classifying each candidate into one of a plurality of predefined categories based upon the candidate's keyword relationship rating, the categories defining the likelihood of the candidate being suitable for the job role. The JWC, table 30 defines what KRW % derived from the KRW table 20 is considered as Class A, B-or C; A being the best fit (most qualified) and C being the least fit (least qualified). Depending on the matching between the KRW % and the JWC definition, a candidate will be classified into Class A, B, C 31 or not classified at all if his/her KRW% is below the cut-off mark of Class C. The cut-off scores for classes A, B, C are variable and may be dependent on the recruiter. Preferably the cut-off scores will be Class A: 70% or above, Class B, 60-69%, Class C, 50-59%.

Take for example, a Candidate X1 who has Job 1 KRW% of 70%. Candidate X1 will be classified as a Job 1 Class A candidate. And for Candidate X2 and X3 who have Job1 KRW% of 80% and 90% respectively, they will also be classified as Job 1 Class A candidates.

At the same time, since Candidate X2 has also accumulated Job 2 KRW% of

80%, he is also classified as a Job 2. Class B candidate. But for Candidates X1 and X3, neither of whom accumulated Job 2 KRW% of equal or more than 50% (the passing weightage % for Job 2 Class C), they are not assigned to any classes for Job 2.

The Job-Candidate Classification data is then stored in the JC table per component 3.

Component 3: Job Candidate Classification (JC) Table

The JC table 40 contains all the matched candidates in each job Class A, B and C, referenced by a candidate number (41 ). The JC table is the ultimate outcome of the system and will be used to list the candidates in each classification to the recruiters. In the JC table 40 shown in figure 3, the table 40 includes the job number (1 , 2) 42, each class (A, B, C) associated with the job numbers 43, and the list of candidate numbers (e.g. 55, 89, 99 for Job 1 , Class A) associated with the job number and the relevant class 44.

Each candidate can only be classified into at most one class of each job role/function.

The relationship between each candidate to the potential job roles/functions can be 1 to x, where x is an integer and inclusive of 0. This means that a candidate has the potential of holding more than 1 job role type in the system.

Recruiters may choose to further filter from the list provided in the JC table by adding new keywords and their relationships.

In another embodiment, the system may undertake additional profiling to group 'similar' candidates to others, even though those 'similar' candidates may not be identified in the JC table in a traditional job role search. A recruiter may be offered a choice to click on 'Search for Similar Candidates' button if they wish to also view 'similar' candidates to a particular shortlisted/referenced candidate. Similar candidates are defined in the system as those having at least 2 similar roles to a reference candidate identified in a JC table. In one example, a Candidate 55 may fall into class A/B/C for each of Job 1 , Job 2, Job 3, Job 4 and Job 5. A 'similar' candidate 66 may be classified as one that calls into class A B/C for only Job 2, Job 3, Job 4 and Job 5. In the traditional job role search for Job 1 , Candidate 66 would not be identified as they do not have a A, B or C score for Job 1. However, as Candidate 66 appears to have many similar roles to Candidate 55 he/she may still be of interest to a recruiter and so will also be identified to the recruiter if the 'Search for Similar Candidates' button is clicked with referenced to Candidate 55. The sorting ^' order will be in the order where similar candidates with the most similar job roles be listed at the top.

The profiling system provides many advantages, including:

1. Intelligently reviewing each registered candidates' resume and classifying the candidate into all the key potential job roles/functions based purely on the suitability to do each job role/function, thereby reducing the time in reviewing resumes and selecting the desired candidates.

2. Eliminating the need for the recruiters to provide the list of keywords and their Boolean relationships, thereby simplifying the search process and criteria.

3. As the candidates have been pre-classified, the system allows for smart, accurate quick profiling of each candidate to the recruiters at the time it is required.

4. Including an additional search function allows recruiters to . further refine their search criteria if they choose to do so over the profiled list of candidates.

Reference throughout this specification to "one embodiment" or "an embodiment" means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the present invention. Thus, the appearance of the phrases "in one embodiment" or "in an embodiment" in various places throughout this specification are not necessarily all referring to the same embodiment.

Furthermore, the particular features, structures, or characteristics may be combined in any suitable manner in one or more combinations. It will be appreciated that persons skilled in the art could implement the present invention in different ways to the one described above, and variations may be produced without departing from its spirit and scope.

Any discussion of documents, devices, acts or knowledge in this specification is included to explain the context of the invention. It should not be taken as an admission that any of the material forms part of the prior art base or the common general knowledge in the relevant art, in any country, on or before the filing date of the patent application to which the present specification pertains.

Claims

CLAIMS:

1. A method for profiling job candidates using searchable information contained in a candidate's resume, the method including:

determining a list of keywords associated with a job role;

using a processor, allocating a keyword relationship rating to a candidate based upon a comparison of information in the candidate's resume to the keywords associated with the job role; and

using a processor, determining candidate classification data by classifying each candidate into one of a plurality of predefined categories based upon the candidate's keyword relationship rating, the categories defining the likelihood of the candidate being suitable for the job role; and

storing the candidate classification data in a database.

2. The method according to claim 1 , wherein allocating the keyword relationship rating includes categorising the keywords into one or more rules, each rule requiring the presence of one or more keywords in the candidate's resume, and based upon said presence of the one or more keywords, allocating a percentage keyword relationship rating for each rule.

3. A method according to claim 2, wherein the total percentage keyword relationship rating for all rules for a particular job role is 100%.

4. The method according to claim 1 , further including the step of providing the candidate classification data and the resumes associated with the candidate classification data to a recruiter.

5. A method for determining potential job candidates for a job role using a processor including:

providing an interface for selecting a job role from a predetermined list of job roles; and

displaying candidate classification data stored in a database, said data comprising details of one or more candidates that have been profiled as potentially suitable for the selected job role, wherein the candidate has been profiled by allocating a keyword relationship rating to a candidate based upon a comparison of information in the candidate's resume to predetermined keywords associated with the job role, and classifying each candidate into one of a plurality of predefined categories based upon the candidate's keyword relationship rating, the categories defining the likelihood of the candidate being suitable for the job role.

6. A method according to claim 5, wherein allocating the keyword relationship rating includes categorising the keywords into one or more rules, each rule requiring the presence of one or more keywords in the candidate's resume, and based upon said presence of the one or more keywords, allocating a percentage keyword relationship rating for each rule.

7. A method according to claim 6, wherein the total percentage keyword relationship rating for all rules for a particular job role is 100%.

8. A system for profiling job candidates using searchable information contained in a candidate's resume including:

determining means for determining a list of keywords associated with a job role;

a processor for allocating a keyword relationship rating to a candidate based upon a comparison of information in the candidate's resume to the keywords associated with the job role, and for determining candidate classification data by classifying each candidate into one of a plurality of predefined categories based upon the candidate's keyword relationship rating, the categories defining the likelihood of the candidate being suitable for the job role; and

a database for storing the candidate classification data.

9. The system of claim 8, further including the processor categorising the keywords into one or more rules, each rule requiring the presence of one or more keywords in the candidate's resume, and based upon said presence of the one or more keywords, allocating a percentage keyword relationship rating for each rule.

10. The system of claim 9, wherein the total percentage keyword relationship - rating for all rules for a particular job role is 100%.

1 1 . The system of claim 8, further including an output display for displaying the candidate classification data and the resumes associated with the candidate classification data to a recruiter.

12. A system for determining potential job candidates for a job role including: an interface for selecting a job role from a predetermined list of job roles;

a display for displaying candidate classification data stored in a database, said data comprising details of one or more candidates that have been profiled as potentially suitable for the selected job role, wherein the candidate has been profiled by a processor allocating a keyword relationship rating to a candidate based upon a comparison of information in the candidate's resume to keywords associated with the job role, and determining candidate, classification data by classifying each candidate into one of a plurality of predefined categories based upon the keyword relationship rating, the categories defining the likelihood of the candidate being suitable for the job role. ,

13. A system according to claim 12 wherein the processor categorises the keywords into one or more rules, each rule requiring the presence of one or more keywords in the candidate's resume, and based upon said presence of the one or more keywords, allocates a percentage keyword relationship rating for each rule.

14. A system according to claim 13, wherein the total percentage keyword relationship rating for all rules for a particular job role is 100%.

15. A system for determining potential job candidates for a job role including: an interface for selecting a job role from a predetermined list of job roles; a display for displaying similar candidate data stored in a database, said similar candidate data comprising details of one or more candidates that have been profiled by a processor as a candidate similar to one or more referenced candidates, said referenced candidates comprising candidates who have been profiled as potentially suitable for the job role using the system of claim 12, wherein said similar candidate data is determined by the similar candidate matching at least two job roles of said reference candidate. ·