Adapting Markov Decision Process for Search Result Diversification

Training data

Files in the data fold

• query_permutation: documents information related to each query

• query_representation: embeddings of queries

• document_representation: embeddings of documents

• query_doc: subtopics covered by each document

Command Line

python RL4SRD.py query_permutation.json query_representation.dat document_representation.dat query_doc.json folder

Training results

The results are saved in folder

Reference

[1] Long Xia, Jun Xu, Yanyan Lan, Jiafeng Guo, Wei Zeng, and Xueqi Cheng. Adapting Markov Decisioin process for Search Result Diversification. In Proc. SIGIR 2017.