Skip to content

Instantly share code, notes, and snippets.

@brh55
Last active May 23, 2016 18:22
Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save brh55/150567b81c63839c3b44c4c362587cf1 to your computer and use it in GitHub Desktop.
Save brh55/150567b81c63839c3b44c4c362587cf1 to your computer and use it in GitHub Desktop.
Notes for in Information Retrieval Systems at Drexel University

The past:

Relevance was always based on subject matter experts.

However...

We measure based on the relevance of the query, despite the fact that other documents may be relevant. It needs to pertain to the query -- that's what we care about.

Contignecy table:

relevant non-relevant
retrieved A & B ~A & B
not retrieved A & ~B ~A & ~B

The Fall Out measure

Essentially getting a measure of non-relevant retrieved items, out of the all non-relevant. Is a good indicator of things that aren't relevant.

Total of Non Relevant and Retrived / Non Relevant

| Ā  & B |
----------
| Ā |
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment