microblog-demos

Examples of using the 2013 TREC microblog API. Includes a few temporal reranking feature.

Getting Stated

Once you've cloned the repository, build the package with Maven:

$ mvn clean package appassembler:assemble

Appassembler will automatically generate launch scripts for three classes:

• target/appassembler/bin/RunQueriesQL: baseline run
• target/appassembler/bin/RunQueriesRecency: recency prior
• target/appassembler/bin/RunQueriesKernelDensity: kernel density estimation

To automatically generate project files for Eclipse:

$ mvn eclipse:clean
$ mvn eclipse:eclipse

You can then use Eclipse's Import "Existing Projects into Workspace" functionality to import the project.

Invoking Sample Runs

After building, you can run any of the sample programs via somthing like this:

$ ./target/appassembler/bin/RunQueriesQL ./config/params_run.json 

which will run a simple baseline query likelihood retrieval. All runnable programs are in ./target/appassembler/bin/ . Also, all programs take a single argument: a JSON-formatted file that will look something like this:

{
"queries"      :  "./data/topics.microblog2012.txt",
"qrels"        :  "./data/qrels.microblog2012.txt",
"oracle_epochs":  "./data/oracle_epochs.txt",
"use_oracle"   :  false,
"host"         :  "<host_name>",
"port"         :  9090,
"num_results"  :  1000,
"num_rerank"   :  5000,
"group"        :  "<your_group_here>",
"token"        :  "<your_token_here>",
"runtag"       :  "uiucGSLIS_01"
}

Most of these variables are self-explanatory. A few aren't. For instance qrel_times is used only for an oracle condition experiment and is probably not of interest for most people--just leave it blank. The variable num_rerank specifies the maximum number of documents to consider for re-ranking per query. The program will print no more than num_results per query.

License

Licensed under the Apache License, Version 2.0: http://www.apache.org/licenses/LICENSE-2.0