kXpert Profiling System using Unsupervised Keyword Extraction

Developing an expert system for identifying experts is crucial for organisation especially higher institution. Finding an expert is challenging because some of the researchers are not frequently update their profile, which significantly influenced on potential research collaboration and student intake in post-graduate studies. This impact has indirectly shown the importance of implementing new method in search for experts, with expertise verified by trusted information resources. The kXpert profiling framework is proposed as a design of a new expert finding system using unsupervised keyword extraction algorithm and data filtering. The algorithms used in this system were RAKE, YAKE and Key-BERT, in which KeyBERT is found to be the most accurate among the three algorithms. The system is expected to reduce the time and effort required to engage in discussion with an individual to determine whether he or she is the correct expert in need