Towards A Job Recommender Model: An Architectural-Based Approach

Abstract

The increase in the number of job Websites that have a multiplier effect on the number of online job information, especially in Nigeria is causing information overload in the area of information management for those seeking for a job. There is a need to fetch job information from different job Websites and suggest job information based on the profile of the user. In the architecture presented in this paper, the user profiles was extracted from social networks and job information from job Websites in Nigeria. Jaccard Similarity Coefficient was used to measure the similarity between terms in the profile of the user and job information to make a recommendation of the job for each user. The architecture complies with service-oriented architecture principles and drew on the Jaccard coefficient for similarity measurement, and content-filtering retrieval feature for the recommendation for its formulation. In determining the preset threshold, similarity coefficients of 275 sample of job information were measure against 10 user profile using classification accuracy. The threshold was set to 0.1 and varied up to 1 per job information to determine the point at which the level of correctness was the highest. The result showed that the lowest level of correctness was achieved at 0.1 thresholds while the highest level of correctness was achieved at 0.8 thresholds. The implementation of job Recommender application(s) from this architecture was able to reduce the problem of information overload and asymmetry. Users of the implemented application was also freed from the problem of subscribing to more than one job website.