User profiles based on user reviews extracted from OpinRank dataset

The OpinRank dataset contains hotel reviews and aspect ratings. There are 5 aspects ratings related to hotels: cleanliness, value, service, location and room. The aspect ratings values are on a scale of 1-5. The authors manually created textual queries related to each aspect rating. These queries were based on real queries made by users in popular search engines, so they reflect a natural user query. For example, the query "great location" is related to the aspect rating "location". Given this query, the dataset lists the aspect rating value of each hotel. The rating values are given by users from TripAdvisor when evaluating the hotels they have visited. In essence, the OpinRank dataset contains five hotels aspects, each aspect is related to user queries, and one aspect rating value is related to each hotel. In addition, this dataset contains hotel profiles composed of unlabeled reviews made by users who have visited the respective hotel.OpinRank数据集（OpinRank dataset）包含酒店评论与属性评分（aspect ratings）。该数据集涵盖5项与酒店相关的属性评分，分别为清洁度、性价比、服务、位置与客房。属性评分采用1至5的五级计分制。研究人员手动构建了与每项属性评分对应的文本查询（textual queries），这些查询源自主流搜索引擎中的用户真实搜索请求，因而能够反映自然的用户检索意图。例如，查询“绝佳的位置”对应“位置”这一属性评分。针对每个查询，本数据集会列出对应酒店的该属性评分值，而评分数据来自用户在TripAdvisor上对其入住酒店的评价。本质而言，OpinRank数据集包含五类酒店属性，每类属性均关联用户查询，且每家酒店对应各属性的一项评分值。 此外，本数据集还包含由入住过对应酒店的用户所撰写的未标注评论组成的酒店档案。