EXTRA

数据集EXTRA由香港浸会大学创建，包含三个子数据集，旨在通过标准排名指标评估推荐解释的可解释性。这些数据集基于用户生成的评论，通过高效的方法如局部敏感哈希（LSH）处理，以确保解释的质量和多样性。数据集适用于推荐系统领域，特别是解决推荐解释的评估和个性化问题。

The EXTRA dataset, developed by Hong Kong Baptist University, comprises three sub-datasets. It is designed to evaluate the explainability of recommendation explanations using standard ranking metrics. These datasets are constructed based on user-generated comments, and processed via efficient methodologies including Locality-Sensitive Hashing (LSH) to guarantee the quality and diversity of the generated explanations. This dataset is applicable to the field of recommender systems, particularly for resolving the evaluation and personalization issues related to recommendation explanations.

搜集汇总

背景与挑战

背景概述

EXTRA数据集由香港浸会大学构建，包含三个子数据集，基于用户评论并通过局部敏感哈希等技术处理，旨在评估推荐解释的可解释性。它主要用于推荐系统领域，解决解释评估和个性化问题，强调解释的质量和多样性。