Design a batched information retrieval system based on a concept-lattice-like structure
作者：
Highlights：
• Creatively propose the concept of Batched information retrieval, which significantly improves retrieval efficiency in certain scenes.
• Propose a batched information retrieval algorithm based on a concept lattice similar structure called keyword-DAG, to fully reuse the similarity among queries.
• Propose an improved rank algorithm with mathematic proof, which greatly shorten the retrieval time.
• Proposed an overall planning algorithm search for improving the efficiency of search and utilization of memory when the batched size is large.
• Implement this method with the popular open source library Lucene and so is easy to be applied in practice.
摘要
•Creatively propose the concept of Batched information retrieval, which significantly improves retrieval efficiency in certain scenes.•Propose a batched information retrieval algorithm based on a concept lattice similar structure called keyword-DAG, to fully reuse the similarity among queries.•Propose an improved rank algorithm with mathematic proof, which greatly shorten the retrieval time.•Proposed an overall planning algorithm search for improving the efficiency of search and utilization of memory when the batched size is large.•Implement this method with the popular open source library Lucene and so is easy to be applied in practice.
论文关键词：Information retrieval,Information filtering,Concept lattice,Inverted file,Ranking algorithm,Caching
论文评审过程：Received 2 August 2017, Revised 24 February 2018, Accepted 27 February 2018, Available online 7 March 2018, Version of Record 26 May 2018.
论文官网地址：https://doi.org/10.1016/j.knosys.2018.02.041
