利用信任及非信任網路之新進使用者商品推薦演算法

Abstract

冷開機的推薦是很重要的，因為它可以幫助建立使用者忠誠度，而使用者忠誠度是電子服務以及電子商務系統的關鍵。推薦有用的資訊跟新的使用者通常可以給他們一種歸屬感進而鼓勵他們常造訪電子商務系統。然而，新的使用者需要時間來熟悉推薦系統，因此系統能擁有的資訊有限而難以產生適合的推薦。這個所謂的冷開機現象對推薦系統的表現有嚴重的影響。 為了解決這個問題，我們提出了一個針對冷開機的推薦方法，結合了使用者模型、信任網路已及不信任網路來辨識出值得信任的使用者。在著名的Epinions的資料集上進行的實驗證實了我們提出的方法是有效以及有效率的。除此之外，此方法在覆蓋率以及執行時間上也優於著名的推薦方法，而不會顯著的降低推薦的準確度。

Cold start recommendations are important because they help build user loyalty, which is the key to the success of e-services and e-commerce systems. Recommending useful information for new users generally creates a sense of belonging and loyalty, and encourages them to visit e-commerce systems frequently. However, new users require time to become familiar with recommendation systems, so the systems usually have limited information about newcomers and have difficulty providing appropriate recommendations. This so-called cold start phenomenon has a serious impact on the performance of recommendation systems. To address the problem, we propose a cold start recommendation method that integrates trust and distrust networks with a user model to identify trustworthy users. The suggestions of those users are then aggregated to provide useful recommendations for cold start users. Experiments based on the well-known Epinions dataset demonstrate that the proposed method is effective and efficient. Moreover, it outperforms well-known recommendation methods in terms of the coverage rate and execution time, without a significant reduction in the precision of the recommendations.