社群事件驅動之適地性商務推薦機制

Abstract

隨著智慧型手機以及社群平台的崛起，行動商務市場逐漸聚焦在"社群"、"適地性"與"行動"三大元素的結合(SoLoMo)。過去的SoLoMo-based的推薦系統透過個人喜好(personal preference)與社群影響力(social influence)來進行廣告、地點或是折價券的推薦，但多數的系統並沒有考慮到使用者當下的"情境(context)"，導致使用者接收到過多且冗餘的推薦資訊，不僅增加使用者的購物成本，也降低適地性商務的商業價值。為了解決上述問題，本研究提出了社群事件驅動之適地性商家推薦機制，透過社群平台中的打卡功能，偵測使用者附近可能正在發生的事件(event)，將此元素加入商家推薦系統中，藉以增強推薦系統的精準度，使得適地性商務的商業價值得到進一步的提升。

With the rise of the smartphone alone with social network platforms, emerging So-Lo-Mo (social-local-mobile) based services bring new business opportunities for mobile commerce. Check-in functionality provided by most of social media play a key role in connecting the online world to the offline world. Many business applications use the check-in information to provide location based service through the recommendation of nearby place, coupon or activities. In this research, a social-event-driven stores recommendation mechanism is proposed. We aim at discovering the possible event the user is attending to identify the possible reason of the users being at that location. Apart from social events, we also consider personal preference and social influence factor as our components in our mechanism. By the synthesis of these components, we are able to significantly enhance the effectiveness of the store recommendation and further improve business values of the social media platform from providing location based commerce.