Information Extraction from Social Media

Abstract

With the proliferation of social media sites, such as Twitter, Facebook, and LinkedIn, social streams have proven to contain the most up-to-date information on current events. Therefore, it is crucial to extract activities or events from the social streams, such as tweets and it become an ongoing research trend. Most approaches that aim at extracting event information from twitter typically use the context of messages. However, exploiting the location information of georeferenced messages and the profile data are also important because tweet messages are short, fragmented and noisy, and therefore not include complete information about the events. For this, in this paper, a framework for event-extraction and categorization from Twitter is proposed. To extract the localized related activities, several mining mechanisms and cleaning techniques is used for real-time twitter corpus and various language processing approaches is applied for categorization the events and then the system will display the valuable information for the targeted domain.