Internet has become the primary medium for Human Resource Management, specifically job recruitment and employment process. Most classical job recruitment portals on the internet rely solely on the keyword based search technique in plain text to locate jobs. However, this technique results in high recall with low precision and also without considering the semantic similarity between these keywords.

Many researchers have also proposed several semantic matching approaches by developing ontologies as a reference to determine matching accuracy qualitatively, however these approaches do not quantify how closely matched applicants and employers are based on core skills.

This dissertation proposes a technique that uses an ontology based approach to enhance keyword searching by leveraging on the similarity between concepts in the ontology, which represent core skills needed and required for a job in order to determine how closely matched an applicant is to a job advertisement and vice-versa.

This was achieved by developing a Curriculum Vitae (CV) Ontology, annotating applicant profile and job postings using a common vocabulary and modifying the semantic concept similarity algorithm to accurately compute and rank matching score between profiles when a query is performed.

The model was compared with the work of Tran (2016). The results showed that improvements were achieved in overall matching accuracy between core skills supplied by applicants and those required by employers. Improvements of 54% and 36% were obtained for Recall and F-measure respectively, over Tran (2016).



1.1 Research Background

Internet has become the primary medium for recruitment and employment processes. According to Jobberman’s Online Recruitment Service Report (2015) (Nigeria’s foremost online job portal), applications on its job recruitment portal increased by over 50% between May and September 2015. This clearly indicates an upward trajectory in online job portals being a major player in contemporary job recruitment process. The relevance of the Internet in job recruitment process cannot be overemphasized, more than three-quarter of the age class qualified for recruitment are active internet users and there is an increasing number of companies that publish their job vacancies on the web.(Report, 2015)

There is a large number of online commercial job portals competing to publish job postings for a fee. On the other hand, each company can publish job postings on its company’s own Website (Mulder, 2010). However, publishing postings on the corporate website reaches a very limited audience, because the indexing capabilities of current search engines are too imprecise to support searches for open positions. Beside this, meta-search engines are limited in their ability to generate offers that match the precise needs of the clients since job postings are written in free text form using uncontrolled vocabulary (Mochol et al., 2007). Furthermore, some dedicated search engines are entering into the market, allowing detailed queries as opposed to keyword-based search of current search engines. However, the quality of search results depends not only on the search and index methods applied, but on the relevance of the search result to the user’s query. Influential factors include the process ability of the used web technologies and the quality of the automated interpretation of the company specific terms occurring in the job descriptions.

The problems of a website’s machine process ability result from the inability of classical web technologies to semantically annotate the content of a given website.Semantic Web extends the Web with machine-understandable data, in addition to classic Hypertext Mark- up Language (HTML) pages. This implies that viable improvements can be made for all parties involved in the recruitment process by tapping into the capabilities of semantic web technologies based on the semantic annotation of job postings and profiles.

In human resource management, it is often necessary to locate and match individuals and positions. Examples of such tasks include human resource recruiting, selecting individuals for teams based on different skills and qualifications, and finding the right expert to acquire information or to learn from within an organization. For human resource recruiting, the Internet is being mainly used to place online job advertisements, to perform resume search, and to acquire information about skills and competencies of individuals (Dafoulas et al., 2003). In order to augment and assist this process, the study and development of totally or partially automated techniques and tools have received the attention of both researchers and organizations. To effectively locate and match individuals and positions, within or from outside an organization, it is important to use semantic technology (Coulucci et al., 2003).

Semantic descriptions of job offers and applicant profiles allow for qualitative and quantitative reasoning about matching between available and required skills and competencies which is needed to improve the process of deciding whom to hire and assigning individuals to tasks and teams. Furthermore, semantic descriptions of applicant profiles within an organization helps improve the management of individual skills and competencies of available human resources, and provide a global view of the skills available at the organizational level. The World Wide Web is a fast growing medium of information and services and as such, there is a need for this information and services to be shared between people and applications alike. Ontologies play a major role in supporting the information sharing mechanism by using semantic and extended syntactic interoperability of the web(Coulucci et al., 2003).

1.2 Statement of Research Problem

In most classical search engines and search mechanism adopted by online job recruitment portals, there is a heavy reliance on containment of keywords in free text before search results are returned which may produce a lot of result from a submitted keyword or phrase but many of these results may be irrelevant to user’s need therefore, a user may have to navigate through a large number of results to find a domain specific results.Many researchers have proposed several semantic matching approaches and have developed prototype job portals to effectively match job seekers with corresponding job postings. They achieved this by developing human resource or Curriculum vitae (CV) ontologies using controlled vocabularies to determine how applicants are closely related to job positions advertised. however, there exists outstanding problems which this dissertation seeks to address, such as quantitatively and precisely matching job seekers with available job postings based on semantic similarity between their core skills and competences in relation to the core skills and competences required for the advertised jobs and also ranking the search according to the semantic closeness between applicant profile and job profile relative to their respective core skills set.

1.3 Motivation

Increasing unemployment rate in Nigeria necessitates the need to develop a system that will ease recruitment challenges by bridging the gap between employable candidates with available jobs. Developing an approach that matches job applicant profiles with available job vacancies based on semantic similarity between the core skills and competencies of a job seeker and core skills required for a job increases the suitability of candidates to jobs they apply and vice versa and gives priority to hiring primarily based on technical competence and know-how. Furthermore, to adopt an ontology based approach in job portals in order to gain advantages of semantic web technology. The semantic web technology enables proper integration of knowledge in ontology based applications.

1.4 Aim and Objectives

The aim of this research is to enhance job search in online job portals by using an ontology based concept similarity matching algorithm to improve matching by core skills.

The specific objectives are to:

1. develop a Curriculum Vitae (CV) ontology based on core skills

2. modify the concept similarity matching algorithm

3. implement the modified concept similarity matching algorithm

4. analyze and compare the result of this work with that of Tran (2016).

1.5 Research methodologies

The following methods will be employed to achieve the stated objectives of this research.

1. Review of relevant literature in the field of study.

2. Determining and Enumerating important core skills in the programming languages domain.

3. Defining and modelling classes and class hierarchy for the ontology using Web Ontology Language (OWL).

4. Allocating weight values to the edge between concepts in the ontology using appropriate weight allocation functions.

5. Enabling node routing by computing related weight between nodes in every path.

6. Determining total sum of all pair concept computation for any two Profiles.

7. Results will be compared with work of (Tran, 2016) in terms of Precision, Recall and F-measure.

1.6 Dissertation Organization

The organization of the rest of this dissertation and a brief outline of the chapters are as follows: Chapter 2 discusses the concept of Human Resource Management on the web and how ontologies have been used to improve it and also reviews related literature. In Chapter 3, the proposed system architecture and modified concept similarity matching algorithm were presented. Chapter 4 shows results obtained from the implementation of the algorithm proposed in chapter 3 as well as evaluation of the results obtained. Chapter five gives a conclusion, summary and recommendation for future studies.

Get Full Work

Use this article as a guide for your own research paper (if properly cited)