Fix io.anserini.ltr.FeatureExtractorCli for argument -collection clueweb #689
Conversation
| if ("clueweb".equals(collection)) { | ||
| return new WebxmlTopicReader(Paths.get(topicsFile)); | ||
| } | ||
| else if ("gov2".equals(collection)){ |
There was a problem hiding this comment.
we use two space indent, and else with preceding braces.
|
Thanks for your contribution! Other than minor formatting changes, we can merge. |
|
@mam10eks do we have an update on how this is going? |
|
Hi @lintool, I am still working on this. Since I am working on this in my spare time I think I need two more weeks until I can finish this and have everything properly tested. My goal is to reproduce some results reported in the "Learning to Rank for Information Retrieval" book (Tie-Yan Liu) with the Anserini-LTR-Pipeline to be sure that everything is fine. |
|
@mam10eks How is the progress going? Are you still working on it? |
|
Hi @MathBunny, I had planned too much and there have been some changes in the Master Branch since then. |
|
From my side this can be merged now. There are still other issues in the LTR-Pipeline. |
|
hi @MathBunny can you take a look? you're more familiar with the code than I am at this point... do a CR and if it's okay we'll merge. |
Following castorini#692 - which fixes castorini#689 - changes need to be propagated to the documentation also.
This pull request will close #688.
I would be happy if I could improve the robustness of
io.anserini.ltr.FeatureExtractorClifurther by adding some integration-tests.If there is some interest in that just let me know, I would be glad to help.