How is the Score in Resume Search calculated?
The score is based on the frequency of the keyword appearing in the text compared to the total number of words in the document.
How can a CV with one relevant word can have a higher score than a CV with multiple words?
For example,:
“I’ve done a resume search on AutoCAD selecting (resumes only and show only highest scored document for candidate).
Candidates with AutoCAD on their CV only once are showing a higher score than candidates with AutoCAD on their CVs multiple times?”
The score is based on the frequency of the keyword appearing in the text compared to the total number of words in the document.
The lower scored document is much larger in terms of the total number of words in the document then the higher scored document.
In your example, the ‘frequency’ of matches within the document is lower because although it has 3 occurrences of the word ‘AutoCAD’ it is 3 within 1000 words whereas the other one has 1 within around 100 words.
Think of it this way; if you have a page of text where AutoCAD is mentioned once, should this score higher or lower against a book (of 100 pages) of info where AutoCAD is mentioned 3 times?
Generally in search theory the single page is more important than the whole book, as the page would seem to be more relevant to the word then the whole book.
In the example mentioned, the score of the higher is 0.647 and the lower is 0.643 – so there is only a very slight difference in score.