Quality Evaluation of Wikipedia Articles / ウィキペディア記事の品質評価

Wikipedia is well known as a free encyclopedia, which is a type of collaborative repository system that allows the viewer to create and edit articles directly in the web browser.
The weakness of the Wikipedia system is the possibility of manipulation and vandalism cannot be ruled out, so that the quality of any given Wikipedia article is not guaranteed. It is important to establish a quality evaluation method to help users decide how much they should trust an article in Wikipedia. We investigate the edit history of Wikipedia articles and propose a model of the network structure between editors. We designed an algorithm to calculate a network structural indicator called restoreratio, which captures percentage of opposing edit actions. We use the proposed indicator combined with existing metrics to predict the quality of Wikipedia articles through a support vector machine. Our experimental results show that the proposed indicator has better performance in quality evaluation than several existing metrics.

Wikipediaは世界最大のオンライン百科事典であり,その特色のほとつは利用者 が共同して記事の執筆および編集を行うことができることである.一方 Wikipediaは誤りやいたずらが混入する可能性があるといった信頼性の問題があ り,Wikipediaの個々の記事は十分な信頼性を持つとは言い難い.個々の記事が どれくらい信頼できるかの情報を利用者に提供することが重要である.本研究で は,Wikipediaの各記事の編集履歴に着目し,参加している編集者間の関係を ネットワーク構造としてモデル化し,対立的な編集行動を行っている割合を示す restoreratioという指標を求める.この新たな指標と既存の指標を組み合わせ, サポートベクターマシンで学習を行うことにより,記事の品質を評価することが できる.実験結果により従来の手法よりも精度の良い評価結果が得られた.
editnetwork.gif

About this Entry

This page contains a single entry by webadmin published on June 9, 2011 2:08 PM.

Coreference Analysis on XML Documents Utilizing Structural Features / 文書構造を考慮したXML文書の照応解析 was the previous entry in this blog.

Recommending Privacy Setttings for SNS / SNSにおけるプライバシー設定の推薦 is the next entry in this blog.

Find recent content on the main index or look in the archives to find all content.