For example, it would be counterintuitive to postulate $Prop: CN$, which is to say that common nouns have propositions as objects. But in the case above, this doesn't seem to work. Matching tokens (i.e skills) should be straightforward, but are there any viable semantic approaches ? I'd appreciate any pointers, as I'm very much new to NLP.Īs far as I know, universes are usually "nested", with $Prop$ being the "first level" (see this answer, for example). That leaves me with similarity-based methods, as far as I can tell. To the best of my knowledge, there is no standard large open source corpus for the person-job fit task, so it seems like supervised methods are off the table. I'm at a bit of a loss as to what approach I should be taking. The parsed dataset features include education, job history (job title and description), and skills. However, It seems to me that tf-idf does not take full advantage of the semi-structured dataset of 1.4k entries that is available to me in JSON format. I have inquired about vectorization and matching methods such as tf-idf and cosine similarity, and those seem to do the job decently. Yansu Asks: How to build a Resume-to-Job Description matcher based on a parsed JSON Resume dataset?įor my capstone project/internship I'm working on an "HR assistant" tool designed to help match, score and rank resumes given a job description and/or requirements.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |