Jimmy Lin

icon indicating a website link https://cs.uwaterloo.ca/~jimmylin/

icon company University of Waterloo icon location Nearby data lake icon location I profess to know very little at the University of Waterloo. I used to write code for Twitter and slides for Cloudera.

Results 265 comments of Jimmy Lin

JsonVectorCollection weights are not obeyed for long terms

sure! Issue noted and PR welcome - but this is lowish on our priority list to fix...

Indexing fails for acl anthology

Can you isolate the troublesome record?

Indexing fails for acl anthology

So, if the error comes from a 3rd party lib, we should just eat the exception and move on? Can we build this into a test case?

BM25 + RM3

> 1. How is p(M_d|q) computed? That is, how does a document's weight in the RM1 model is defined when the initial retrieval is BM25? https://github.com/castorini/anserini/blob/master/src/main/java/io/anserini/rerank/lib/Rm3Reranker.java#L83 > 2. After receiving...

mvn error

Sometimes this is a transient error... can you try again?

Create the RankingModel class to wrap up the logical ranking model

Ref #1038 - initial steps in that PR.

Integrate Waterloo spam scores and other static priors into index

Hrm... that's pretty heavyweight and requires an external dependency. I suppose for catB everything can fit in memory. Perhaps we can assume the same for catA? 500m \* ( 2...

Integrate Waterloo spam scores and other static priors into index

I'd rather index everything and use spam as a feature during retrieval. That way we don't need to develop a cutoff.

Integrate Waterloo spam scores and other static priors into index

Just a big hashmap we load into memory at startup? Using, fastuil, for example?

Integrate Waterloo spam scores and other static priors into index

How much memory do you have on your machine? The machine I use at UMD has 0.75 TB RAM :)