A heuristic hierarchical scheme for academic search and retrieval

作者：

Highlights：

• We present PubSearch, a new system for academic search and retrieval.

• PubSearch is based on a cascade hierarchy of three heuristics.

• The heuristics include Term-Frequency, citation distribution and topics’ inter-relations.

• We compare PubSearch performance against ACM Portal on 58 user queries.

• The system outperforms ACM Portal in terms of ERR, NDCG & LEX metric by large margin.

摘要

•We present PubSearch, a new system for academic search and retrieval.•PubSearch is based on a cascade hierarchy of three heuristics.•The heuristics include Term-Frequency, citation distribution and topics’ inter-relations.•We compare PubSearch performance against ACM Portal on 58 user queries.•The system outperforms ACM Portal in terms of ERR, NDCG & LEX metric by large margin.

论文关键词：Academic search,Search and retrieval,Heuristic document re-ranking

论文评审过程：Received 21 December 2011, Revised 25 June 2013, Accepted 1 July 2013, Available online 31 July 2013.

论文官网地址：https://doi.org/10.1016/j.ipm.2013.07.002