Using Big Data Technique for Building Edit Alert System for Wikipedia Infoboxes Based on Map Reduce Method

Abstract

Wikipedia is an online encyclopedia and has  become a vital information resource for users as well as  for many knowledge bases derived from it. This  information requires manual editing for update. Wikipedia  provides an infobox on the right hand side of many  articles. An infobox of a Wikipedia article generally  contains key facts in thearticle and is organized as  attribute-value pairs. All the Wikipedia’s content is  manually updated or maintained by contributors. This  leads to the fact that its information is not updated  regularly and completely. In this paper, we present a novel  system that focuses onprediction of data items that are  most likely to be updated, based on the category of page,  record key, last time updated, etc. for alerting Wikipedia  editors, about the data items that might need update soon,  using Time series modeling. Concept of Bipartite graph is  used to perform user based collaborative filtering to find  similar editors who might be interested in editing the  infobox. The update alert is sent to editors found using  Bipartite graph along with the past editors of a particular  infobox. The technique to deal with vandalic and  erroneous edits is also discussed and its analysis is given. We have also presented various tasks that can be carried  out on infoboxes.