Diversifying citation recommendation by combining semantics and time of academic papers

Zhang, Xiaojuan

Permalink

https://hdl.handle.net/2142/106578 Copy

Description

Title

Diversifying citation recommendation by combining semantics and time of academic papers

Author(s)

Zhang, Xiaojuan

Issue Date

2020-03-23

Keyword(s)

• Citation recommendation
• Diversification
• Publication time
• Semantics

Abstract

In this paper, we present a study of a novel problem, i.e., diversifying citation recommendation by combining the semantics and publication time of academic papers, and we seek to generate a list of rec- ommended citations that consist of relevant ones covering diverse semantics, and at the same time, being related to different time periods. Two major tasks are involved in our work. In the first task, a unified graph model is used to generate candidate citations for each query manuscript. In the second task, candidate citation diversification is carried out by combining semantics and time of academic papers in an implicit diversification way. Preliminary experiment on ANN dataset demonstrates that our proposed method beats the baselines in terms of metrics used in citation recommendation ranking and diversification.

Publisher

iSchools

Series/Report Name or Number

iConference 2020 Proceedings

Type of Resource

• text
• image

Language

eng

Permalink

http://hdl.handle.net/2142/106578

Copyright and License Information

Copyright 2020 Xiaojuan Zhang

Owning Collections