Adding MMR and Portfolio IR rank re-adjusting #2
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Both batch and evaluation mode command line invocation is possible. For further information pass ‘—help’.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Search result diversity re-ranking and evaluation - submission for the IRDM group project.
Group 22: Abhishek Aggarwal, Clive Austin, John Lyon
This is a submission of code for our group project of COMPGI15 Information Retrieval and Data Mining. We forked the original repository and added two features onto the Panda platform:
The -MMR or -Portfolio mode : These modes readjusts ranking.
The -bMMR or -bPortfolio mode: These can run batch evaluation on MMR/Portfolio hyper-parameters.
See Panda Help (Panda.java)
Examples command line options:
Evaluation
-MMR 10 -m 2: readjusts top 10 ranks using MMR algorithm on results retrieved using BM25 model (#2). Default value of lambda = 0.5 as found optimal. See report. This default can be changed.
-Portfolio 20 -m 2: readjusts top 20 ranks using Portfolio IR algorithm on results retrieved using BM25 model (#2). Default value of hyper-parameter b = -1 as optimal. See report. This default value can be changed.
Batch Evaluation
-bPIR 10:-4:0.5:4 -m 2: readjusts top 10 documents using Portfolio IR, with underlying model as BM25 in
batch mode. That is, for the hyper parameter (b) taking value from -4 to +4 with
increments of 0.5
-bMMR 30:0:0.1:1 -m 2: readjusts top 30 documents using MMR, with underlying model as BM25 in batch mode. That is, for the hyper parameter (lambda) taking value from 0 to 1 with increments of 0.1
The implementation is well commented as well as the results and methodology documented in the report submitted. For any questions, please email - [email protected]