Currently the resulting job listings are tagged with a relevance score, which is relative to the pool of results fetched. This gives the user little indication regarding actual relevance. One idea is to adjust the scoring range to fall between something ideally relevant (e.g., practically identical to the user's profile) and something completely irrelevant (e.g., gibberish text). Also, this range can then be divided discretely to tags: High, Medium, and Low. This may even allow filtering out very low-relevance results altogether.
Currently the resulting job listings are tagged with a relevance score, which is relative to the pool of results fetched. This gives the user little indication regarding actual relevance. One idea is to adjust the scoring range to fall between something ideally relevant (e.g., practically identical to the user's profile) and something completely irrelevant (e.g., gibberish text). Also, this range can then be divided discretely to tags: High, Medium, and Low. This may even allow filtering out very low-relevance results altogether.