運用不同觀影設備及不同觀影偏好對OTT平台內容推薦系統的體驗差異影響之研究：以Netflix為例

Abstract

隨著OTT（Over-the-Top）影音平台的崛起，Netflix作為全球最大的OTT影音平台之一，已經成為人們娛樂生活中不可或缺的一部分。在Netflix平台上，推薦系統扮演著非常重要的角色，它可以根據用戶的個人偏好和觀看歷史等資訊，向用戶推薦最符合其喜好的影視節目。而現有的研究大多僅限於探討Netflix的推薦系統模型，缺少對實際推薦結果的評價和用戶的反應的探討。因此本研究旨在探討不同觀影設備和偏好對推薦系統體驗的影響，並分析其與觀影動機和滿意度的關係，採用混合研究方法，量化部分透過便利抽樣收集401份有效問卷，並進行五位年齡介於20至40歲受訪者的深度訪談。量化分析結果顯示，男性用戶和每周觀看時間超過10小時的用戶具有較高的使用動機，而從未使用過Netflix的受訪者對該平台表現出極大興趣。平板電腦使用者的使用動機也相對較高。訪談結果顯示，受訪者普遍偏好使用手機和電視觀看影片，對平台介面表示滿意，但希望推薦系統能更精確匹配其偏好。觀影偏好多樣性和設備選擇主要依賴於使用情境和便利性。本研究建議OTT平台提升推薦系統的準確性和個性化，以改善用戶體驗和增強平台競爭力。未來研究可進一步探討不同用戶群體的觀影行為及跨平台推薦系統的效能比較。With the rise of OTT (Over-the-Top) media platforms, Netflix has become an essential part of entertainment in people's lives, emerging as one of the largest OTT platforms globally. The recommendation system on Netflix plays a crucial role by suggesting shows and movies that align with users' personal preferences and viewing history. While existing research often focuses on Netflix's recommendation system models, there is a lack of studies evaluating actual recommendation outcomes and user reactions. This study aims to investigate the impact of different viewing devices and preferences on the recommendation system experience and analyze its relationship with viewing motivation and satisfaction. A mixed-method approach was adopted, collecting 401 valid questionnaires through convenience sampling for the quantitative analysis and conducting in-depth interviews with five respondents aged 20 to 40 for the qualitative analysis.The quantitative analysis revealed that male users and those who watch more than 10 hours per week have higher usage motivation. Additionally, respondents who had never used Netflix showed a strong interest in the platform, and tablet users also demonstrated relatively high usage motivation. The interviews indicated that respondents generally preferred using smartphones and TVs for viewing, expressed satisfaction with the platform's interface, but desired more precise recommendations. Viewing preferences and device choices were primarily influenced by context and convenience. This study suggests that OTT platforms should improve the accuracy and personalization of their recommendation systems to enhance user experience and strengthen competitiveness. Future research should further explore the viewing behaviors of different user groups and compare the effectiveness of recommendation systems across platforms.