問答網站社群之互補問答知識文件推薦

Abstract

隨著網路科技及Web2.0概念的蓬勃發展，問答網站逐漸成為重要的知識分享平台。問答網站提供知識社群的服務機制，讓擁有共同興趣或專長的使用者組成知識社群。社群中成員能收藏有興趣之問答知識，並分享與社群相關的知識議題。 然而問答網站每天有大量的問答知識產生，造成了資訊過量的問題，因此社群知識收藏之推薦機制應運而生，用以推薦知識社群相關有興趣之問答知識。然目前相關文獻少有針對問答網站社群知識收藏的群體推薦機制之研究。而傳統群體推薦機制多是以群體成員之重要性作為權重，結合各單獨成員之興趣特徵檔以產生群體興趣特徵檔，進而以群體興趣特徵檔過濾推薦物件，並未考量推薦物件如問答知識之品質、知識文件之相關互補性，以及社群成員收藏知識之聲望等因素。 本研究提出問答網站社群知識收藏之群體推薦機制，以推薦社群相關且有興趣的問答知識文件。所提的推薦方法主要以社群中歷史收藏知識之推薦分數、收藏時間及知識成員之重要性包括收藏知識聲望與回答知識聲望等，再根據不同的知識主題產生社群群體興趣特徵檔，並考量知識文件之間的相關互補性，與知識文件之品質，進而推薦具品質之社群相關互補問答文件集，以滿足社群成員對於問答知識的需求，促進知識分享的交流。 最後本研究以奇摩知識家問答網站做為實驗評估的資料來源，實驗結果顯示本研究所提出的方法比傳統方法能更有效的針對知識社群推薦與其興趣相關的知識文件。

With the rapid development of Internet and Web 2.0 technology, Question & Answering (Q&A) websites have become an essential knowledge sharing platform. This platform provides knowledge community services where allows users with common interests or expertise to form a knowledge community. Community members can collect and share Q&A knowledge (documents) of their interests. However, due to the massive amount of Q&A documents created every day, information overload become a main problems. Consequently, a group-based recommendation mechanism is needed to recommend Q&A documents for communities of Q&A websites. Existing studies did not investigate the recommendation mechanisms for knowledge collections in communities of Q&A Websites. Traditional group-based recommendation methods use member importance as weight to consolidate individual profiles and generate group profiles, which in turn are used to filter out items of recommendation, but do not consider certain factors of recommended items, such as the quality of documents, the reputation of community members, and the complementary relationships between documents. In this study, we will propose novel group-based recommendation methods to recommend related Q&A documents for knowledge communities of Q&A websites. The proposed recommendation method build several community topic profiles by considering factors such as community members’ reputations in collecting and answering Q&A documents, push (recommendation) scores of Q&A documents and the collected time of Q&A document from the historical collected Q&A documents, and make recommendations via considering the quality of Q&A documents and their relevance to the communities. Moreover, we further investigate the methods for analyzing and recommending complementary Q&A document sets to satisfy community members’ knowledge needs and facilitate knowledge sharing in communities. This research will evaluate and compare the proposed methods by using the experimental dataset collected from the Yahoo! Answers Taiwan website. Experimental results show that the proposed method outperforms other conventional methods, providing a more effective manner in recommending Q&A documents to knowledge communities.