Trace-based contextual recommendations
作者：
Highlights：
• User interactions are used to build contextual recommendation system.
• A Trace-Based Reasoning process is applied to retrieve and adapt similar episodes.
• An evaluation protocol measures the performance, accuracy and user satisfaction.
• Similarity measures are used to compare between interaction traces.
• The Scenario example is a specific video composition application called Wanaclip.
摘要
•User interactions are used to build contextual recommendation system.•A Trace-Based Reasoning process is applied to retrieve and adapt similar episodes.•An evaluation protocol measures the performance, accuracy and user satisfaction.•Similarity measures are used to compare between interaction traces.•The Scenario example is a specific video composition application called Wanaclip.
论文关键词：Recommender systems,Trace based reasoning,Human computer interaction,User assistance,Interaction trace,Similarity measure
论文评审过程：Received 30 August 2015, Revised 15 May 2016, Accepted 22 July 2016, Available online 27 July 2016, Version of Record 1 August 2016.
论文官网地址：https://doi.org/10.1016/j.eswa.2016.07.035
