由於旅館社群平台的興起,如何從各種異質性資料中找出有趣的關聯性,以及如何利用這些關聯性幫助使用者成為重要的議題。在本篇論文中,我們基於現有的旅館資料庫,再從旅館社群平台以及電子地圖服務中蒐集了更多資料以及影像資訊進行分析。藉由這個多樣性資料庫,我們分析不同屬性資料的關係並探討其有趣的特性。其中,我們發現在旅館評價中,不同文化中的使用者有不同的評價模式。此外,我們更跳脫傳統文字資料分析的範疇,利用視覺分析技術瞭解旅館封面照片,並探討視覺資訊與旅館評價間的關係。我們最後將這些關聯性應用在改善旅館推薦系統,並從實驗中證明加入視覺資訊和文化差異能夠進一步提升預測的效果。
Due to the emergence of hotel social media platforms, how to discover interesting properties and utilize these discovered characteristics in hotel-related applications become important issues. In this thesis, based on a large-scale hotel information collection, we further crawl more hotel information and hotel photos from a web-based hotel social media platform and a web-based map service, in order to conduct advanced analysis. With this rich dataset, we analyze various correlations between hotel properties and interesting characteristics. We found that travelers from different cultural areas (countries) have different rating behaviors. In addition, beyond the scope of conventional text-based hotel analysis, we utilize visual analysis techniques to analyze hotel cover photos, and investigate the relationship between rating behaviors and visual information. We finally adopt these correlations to develop a hotel recommender system, and verify that by considering visual information and cultural difference, recommendation performance can be improved.