Research On Social Media Event Mining And Application

With the development of web2.0, more and more social media begin to enter people’s lives. Social media is a new online media which can give the users a great participation. Microblogging, Wikipedia and Del.icio.us, are examples of social media. Social media is open and it plays a key role in the flow,interaction of information and community formation especially in this era which network information is extraordinary flooding. When an event occurs in our society, a large number of corresponding event-related information will appear on the Internet, such as pictures, videos, micro-blogs, and news reports. Facing this massive picture information in the social media, digging out the event-related pictures is a cumbersome, laborious work which always with a low accuracy. Therefore, how to solve this problem and use a reasonable image tag clustering to retrieval and analysis the image, as well as how to carry on a reasonable and effective clustering on these disorganized image tag have become a hot research topic in the field of social media.The purpose of this paper is to cluster the image tag from the respective side of events therefore we can have a better describe of the event information from the displayed pictures to facilitate event image retrieval. There has been a lot of work and research currently about relationship between the event and the image tag, but for image tag clustering from respective side of the event is still not very mature. In this paper we will have a further exploration with this image tag clustering method we just mentioned above based on previous research findings about events mining, information extraction and image tag clustering. We cluster the image tags by mapping the image tags associated with descriptors collection of each side of the event digging from micro-blog. Finally, we verify the feasibility and effectiveness of this method through our experiments.