ENROLLMENT SELF-SERVICE ASSISTED BY ARTIFICIAL INTELLIGENCE REFERRAL KNOWLEDGE MANAGER ENROLLMENT GAP ANALYSIS MATCH PROFILING SEMANTIC SEARCH FOUNDATION
Ask any PES counselor: helping people to find suitable jobs is hard. There are just so many aspects to consider: age, experience, education, character, company culture, expectations, ambitions, skills, competences, physical location... There is no one size fits all solution. Every jobseeker, employed or unemployed, should be able to get personalized support. Some will only need a little, others may need a lot. Computer systems to help people find suitable work have been around for a while. But where counselors can improvise, apply common sense, and rely on past experience, computers use rules, decision trees, and patterns found in big data. When this input is poor (lacking information, misinterpretation of job titles), the output cannot be trusted. Good results are based on two critical factors: high-quality input and smart processing. The brochure Knowledge Management describes how WCC supports smart processing. This brochure focuses on high-quality input. Enrollment data can come from many sources: 1. a counselor doing a jobseeker intake; 2. a jobseeker using self-service registration; 3. an employer entering a vacancy; 4. a batch process that receives vacancy data. The Challenge The two main problems during enrollment are bad information and missing information. In the first case, people enter information the system cannot use. For example, they use job titles that are not yet in the system s knowledge base. When the system tries to match these job titles to vacancies, nothing happens.
In the second case, people omit important information. A truck driver who forgets to enter his truck driver s license will fail to be matched with any jobs that require such a license. This problem gets even worse in more sophisticated matching processes. For example, skill-based matching only works well when enough meaningful skills are entered. Many jobseekers are unable to do so. But it is not just jobseekers who can cause these issues. PES counsellors doing intakes may use different names, descriptions, and skills than their colleagues. This influences the result without the counselor even realizing. Employers too can become very creative when making vacancy descriptions attractive to jobseekers, but their creativity will only confuse computer systems. Finally, vacancies that arrive in batches (possibly parsed or scraped from websites) can also contain information the system does not yet recognize. Again, these vacancies will be misinterpreted, and the matching process will produce unusable results. The Solution WCC offers two tools to improve the quality of enrollment data: Input Completion and Vacancy Classifier. Input Completion To get good results, people need support when entering job descriptions, skills, diplomas, and so on. That is why WCC offers a powerful software component called Input Completion. This helps people enter high-quality data which the system can interpret correctly and easily. This in turn improves all processes down the line. For a jobseeker registering resume data in self-service mode, Input Completion mimics the intake process normally performed by a counselor. It asks the jobseeker to enter certain data and gives suggestions for improvement. For example, if the system does not recognize the job description entered by the jobseeker, it will suggest alternatives. The system also presents the diplomas, skills, and required licenses normally associated with a specific job title. It is then up to the jobseeker to select the ones that apply. Now the registration information is correct and complete, making a good match much more likely. The advice jobseekers receive is based on the system s knowledge base. Taxonomy and ontology form its basis, but PES may influence the process if desired. Input Completion allows the PES to configure the suggestions made to the jobseeker. For example, it may choose to keep the free text entry of the jobseeker, but use the formal terms behind the scenes. This feature is useful when a PES want to differentiate results for specific user groups, for example by adjusting the language used to address each group. It can also capture new job titles and skills in the market. Chatbot An Artificial Intelligence assistant can also guide jobseekers through enrollment. Let s take the case of Sarah, a social worker looking for a job. Hi Sarah, how can I help you? Hi, I m looking for a job as a caseworker. Can you give me some more information? I have five years experience as a caseworker at a health clinic. I d like to work with children. Do you agree to give me access to your resume data? Yes. Are you registered with a professional organization? I don t see that information in your resume. Yes, I m a member of the National Association of Social Workers. Thank you. I see that there are five skills associated with your current job. You can see them to the right. Do you want to adjust or add any skills? This can help you find more suitable jobs. No, thank you. In that case, let s go to the next step and find you a great job.
Input Completion will also make intakes by counselors much more efficient and of higher quality. For example, skills associated with a job title are automatically presented, and the system makes sure all relevant information is captured in a standard way. Employers benefit from using Input Completion too. When they supply vacancy data, they get suggestions for more commonly used job titles, associated skills, and other related data. That helps them find better fitting job applicants. automatically improve suggestions, for example by ordering them according to how often users selected each suggestion for a taxonomy-registered occupation value. Multilingual input WCC s Input Completion can use multiple languages. Currently there are language packs for English, Spanish, Arabic, and several others. It is also possible to create new language packs. It takes about 10,000 resumes for the system to learn a new language. Vacancy Classifier Vacancy and resume data can be entered manually, but is usually supplied in large batches. That is when WCC s Vacancy Classifier comes in to play. It checks all job titles against the knowledge base and maps them with the official job titles. Later, this information can be enriched with the skills and education associated with these titles. This smart mapping and context-based data enrichment is an example of how WCC uses Artificial Intelligence to improve the quality of the input. An important part of the Vacancy Classifier is monitoring and analysis. The Vacancy Classifier monitors all job titles by logging them. Analysis of this information will provide insight into, for example, changes in the job titles used in the market. That in turn may uncover a need for adjusting the national taxonomy and thus the knowledge of the labor market. The Technology WCC s Input Completion and Vacancy Classifier are powered by the latest in innovative technology. Data-driven Machine Learning WCC s Input Completion is not a static service. It uses data-driven machine learning to learn from what users are doing. When jobseekers make certain choices more often than others, the system will change the order in which suggestions are presented. Users experience higher levels of user-friendliness. Artificial Intelligence WCC s AI-powered knowledge base consists of automatically updated taxonomy data, ontologies, configurable rules, and advanced language technology. These enable Input Completion to guide users through the registration process, asking for information and presenting suggestions in an understandable and personalized way. These suggestions are based on the most recent labor market information powered by Artificial Intelligence. Later on, the same AI also provides personalized digital career guidance and helps jobseekers close gaps to suitable jobs and ambitions. Management Tools The key concept behind Input Completion is mapping word clusters to list values from a standard list (such as a taxonomy describing occupations or educations). Input Completion offers analysis and management tools to control the creation, import, maintenance, and mapping of word clusters. Multiple services can be configured and connected per value type (such as occupation, education, or skill) to relevant data entry fields in the application. All user entries are logged and can be analyzed by functional administrators in Input Completion Management. The resulting information can be used to Semantic Extraction Uploading job and jobseeker profiles is simple but classification of profiles based on topics is often not consistent. WCC s Vacancy Classifier processes profile data in three steps: Semantic extraction: all the relevant information/concepts are extracted Context classification: all the profiles are automatically classified using an extended framework (compliant with O*NET, ISCO, ESCO, SSOC, ASOC, and so on).
User interface Input Completion Ontology enrichment: the classifier adds synonyms, variations, and related concepts to all the extracted concepts. WCC Ontology Job-related data is often ambiguous. Is a software developer the same as a software designer or a software architect? Many terms describe nearly the same thing without being perfect synonyms. When they interact with the PES, nonexperts like jobseekers or even corporate recruiters should be able to use their own vocabulary, instead of being forced to adopt the national occupation classification systems PES consultants and counselors typically use. WCC Ontology solves this problem. It keeps on learning by analyzing millions of job descriptions, jobseeker profiles, and training descriptions. It continuously extracts job titles, skills, and education, finding new synonyms and related concepts. All this information is used to update and enrich the core ontology and thus improve the labor market knowledge base. The WCC Ontology uses advanced AI, machine learning, and ontology creation technology. As one of the largest HR ontologies in the industry, it offers the following standout features: recognizes over a billion concepts includes more than 300,000 jobs/functions and skills/competences covers more than 1,500 job clusters handles over 300 industries is available in over 10 languages
The Features The Benefits Input Completion can: manage word clusters manage Input Completion Services edit service configuration import list values and mappings manage mappings use logs and reports present a list of suggestions based on the input text return suggestions filtered by the context log information about the user s choice WCC s Input Completion and Vacancy Classifier: dramatically improve input quality strongly improve user-friendliness and user satisfaction make registration much faster and easier for end users enable successful implementation of self-service improve the efficiency of the intake process enable advanced digital career guidance offer insights into labor market changes supply information for adjusting and updating the labor market knowledge base improve their own performance automatically
About WCC Our vision People in organizations make decisions. In the markets we focus on, those decisions profoundly impact people s lives. To make the right decisions in an increasingly complex world, it is necessary to have excellent software. That is what drives us at WCC: enabling people to make better decisions. Our mission & strategy WCC wants to give people the answers they need, not just the ones they asked for. We thrive on developing software that can connect, combine, and make sense of large amounts of data stored in different systems. Software that can communicate with the users in a human way, and that delivers superior results so our customers can make a difference. We call this software that matters. But great software alone is not enough to get the best results. What sets WCC apart is the combination of remarkable software with in-depth knowledge of our customers business. That is why business and implementation consultancy is an important part of our strategy. We focus on two markets: Employment and Identity. The security needs of the Identity market are stringent. Border management and law enforcement agencies face the challenge of quickly and accurately identifying people from huge amounts of data spread over many different databases and formats. WCC s software incorporates the necessary evidence-based algorithms, such as multi-cultural name matching, to make correct identifications. HERMES, our API/PNR solution, adheres to industry standards and is easy to implement and operate. Our customers include UNHCR and the European Union. Our products and services The core of the Employment market is matching people with sustainable jobs effectively and efficiently. WCC has proven to be unequalled in doing just that. Our Employment Platform, which combines unique search and match capability with advanced gap analysis and referral to the right measures, delivers superior strategic value to our customers. Many of the world s largest employment and staffing organizations use our products and expertise, including Randstad, Robert Half, and the public employment services of Germany, France, and the Netherlands. WCC Smart Search & Match WCC Services US Inc. Zonnebaan 19 228 Hamilton Avenue 3542 EA Utrecht Suite 300, Palo Alto The Netherlands CA 94301, USA T: +31 (0)30 750 32 00 T: +1-888-922 9224 info@wcc-group.com www.wcc-group.com BLECHQD01