ecir2017-fusion's Introduction

Fusion-Based Object Retrieval

This repository provides resources developed within the following paper:

S. Zhang and K. Balog. Design Patterns for Fusion-Based Object Retrieval. In ECIR'17, April 2017.

This study is an effort aimed at reproducing the result presented in the Fusion-Based Object paper.

This repository is structured as follows:

config/: config files to index data(Using Elastic, ip:port number would change individually)
nordlys/: code required for runnning fusion models
data/: elastic index, query and evaluation files for blog distillatioin(./trecblog/), expert serach(./trecent/) and vertical search(./trecfed/)
lib/trec_eval/: TREC evaluation file
output/: all run files scripts and their result files

Data

The data we used are public data sets:

CSIRO: The dataset was used for TREC task of expert search task in 2006 and 2007.
Blogs06: The dataset was used for TREC task of blog distillation in 2006 and 2007, which is not for free yet.
FedWeb13 and FedWeb14: These datasets were used for TREC task of federated search in 2013 and 2014.

Runs

All the run files can be found in /output. E.g,

python -m output.blog07.early_bm25_run

executes the run of early fusion model incorporating BM25 methods for blog distillation task of 2007.

Qrels

The qrel files are all provided in data/ sub-folders.

Citation

@inproceedings{Zhang:2017:DPF,
    author = {Shuo Zhang and Krisztian Balog},
    title = {Design Patterns for Fusion-Based Object Retrieval},
    booktitle = {Proceedings of the 39th European conference on Advances in Information Retrieval},
    series = {ECIR '17},
    publisher = {Springer},
    pages = {684--690},
	  year = {2017}
}