The Application of Extended Weighted Tree Similarity Algorithm for Similarity Searching

Unemployment is a term given to someone who has not have a job or looking for a job. High unemployment can also affect the economy because it can cause poverty and other social problems. To increase labour force participation, media for searching a job are needed to provide job information and recommendations that are match with job seekers profile.Extended Weighted Tree Algorithms is one of the semantic search algorithms that existing metadata consists of trees, labeled nodes, labeled and weighted branches. The similarity calculation will be done by comparing two tree structures, namely the job advertisement tree with the job search tree carried out by job applicants. The extended weighted tree similarity algorithm is string matching on the leaf node. Thus, if the leaf node is the same then it will produce a value of 1, whereas if it is different it produces a value of 0.The metadata of the Extended Weighted Tree Similarity Algorithm is arranged based on semantic information such as taxonomy, ontology, preference, synonym, homonym, and stemming. The metadata can be used to represented searching results with higher precision than full-text. This research produces an accuracy rate of 88%, precision level of 84.30% and recall of 93.40%.