@techreport{oai:ipsj.ixsq.nii.ac.jp:00054068, author = {松澤, 俊典 and 山口, 未来 and 和田, 雄次 and 土肥紳一 and Syunsuke, Matsuzawa and Mikuru, Yamaguchi and Yuji, Wada and Shinichi, Dohi}, issue = {13(2008-CE-093)}, month = {Feb}, note = {近年,eラーニングが普及し,企業の研修や大学の講義などで取り入れられている.この背景にはSCORMやWebCTなどのLMSの発展が上げられ,システムの運用だけでなくコンテンツの導入も容易となったことにある.コンテンツが増大することは学習の幅を広める一方,学習者がコンテンツ選択の際,何を学べば良いのか迷うといった懸念がある.本研究では,学習者が遷移したコンテンツの順番,つまり学習者がどのように学習を行ったかという学習履歴を基に,コンテンツの関連性を求め,推薦する双方向推薦システムの開発を行った.本稿では,著者らが開発した復習支援システムAIRS(An Individual Reviewing System)の概要及び,双方向推薦システムの詳細と,その評価と有効性について述べる., Recently, e-learning systems have begun to be used in companies and education institutions. The spread of e-learning is mainly owing to development of LMS for instance SCORM and WebCT. Thereby, the introduction of learning texts have became much simpler. The increase of the contents has realized the wider width of learning. As the result, we are afraid that learners are confused with selecting best learning texts. In our research, we are developing bidirectional recommendation system that extracts the relationship among the learning texts with historical logs and recommends the effective text for the learners. In this paper, we firstly discuss the design of the bidirectional recommendation system, secondly show its evaluation results and finally we conclude our recommendation system is useful for the learners.}, title = {教材コンテンツ双方向推薦システムの実装}, year = {2008} }