Matching Resumes and Jobs Based on Relevance Models

摘要

We investigate the difficult problem of matching semi-struc- tured resumes and jobs in a large scale real-world collection. We compare standard approaches to Structured Relevance Models (SRM), an extension of relevance-based language model for modeling and retrieving semi-structured docu- ments. Preliminary experiments show that the SRM ap- proach achieved promising performance and performed bet- ter than typical unstructured relevance models.