1 / 11

Lecture 3: Retrieval Evaluation

Lecture 3: Retrieval Evaluation. Maya Ramanath. Benchmarking IR Systems. Result Quality Data Collection Ex: Archives of the NYTimes Query set Provided by experts, identified from real search logs, etc. Relevance judgements For a given query, is the document relevant?.

tryna
Télécharger la présentation

Lecture 3: Retrieval Evaluation

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Lecture 3: Retrieval Evaluation Maya Ramanath

  2. Benchmarking IR Systems Result Quality • Data Collection • Ex: Archives of the NYTimes • Query set • Provided by experts, identified from real search logs, etc. • Relevance judgements • For a given query, is the document relevant?

  3. Evaluation for Large Collections • Cranfield/TREC paradigm • Pooling of results • A/B testing • Possible for search engines • Crowdsourcing • Let users decide

  4. Precision and Recall • Relevance judgements are binary – “relevant” or “not-relevant”. • Partition the collection into 2 parts. • Precision • Recall Can a search engine guarantee 100% recall?

  5. F-measure • F-Measure: Weighted harmonic mean of Precision and Recall Why use harmonic mean instead of arithmetic mean?

  6. Precision-Recall Curves • Using precision and recall to evaluate ranked retrieval Source: Introduction to Information Retrieval. Manning, Raghavan and Schuetze, 2008

  7. Single measures Precision at k, P@10, P@100, etc. and others…

  8. Graded Relevance – NDCG • Highly relevant documents should have more importance • Higher the rank of a relevant document, more valuable it is to the user

  9. Inter-judge Agreement – Fleiss’ Kappa N– number of results n – number of ratings/result k – number of grades nij – no. of judges who agree that the ithresult should have grade j.

  10. Tests of Statistical Significance • Wilcoxon signed rank test • Student’s paired t-test • …and more

  11. end of module “IR from 20000ft”

More Related