Not in-stock currently-usually arrives within 1-14 business days
Nowadays, social media mining has become a very important part of modern telecommunication systems. It has become essential to find a technique to extract efficiently and effectively the tweets and photos from a large database. Social media sites include Facebook, Google Plus+, Twitter, Linked in, Pinterest, Instagram, Tumblr and Flickr. Every social media websites permit user to publish as well as share the multimedia content which plays a significant role in building of link among users and content (Min-chul et al. 2014). The researchers and academics recently started to examine a variety of social media mining techniques to support experts in enhancing social media. This technique authorizes the experts to detect new information derived from users' data. In this research work, our attention is focused on the Twitter, a micro-blogging social networking website. Twitter generates huge amount of tweets that cannot be handled by hand to extract some useful information and therefore, the ingredients of automatic classification are required to handle the data. Tweets are explicitly short text messages that are up to a limit of 140 characters. Millions of users are connected with their friends, colleagues and family through their computers or mobile phones around the world by Twitter (Saad Ahmed et al. 2017). Twitter interface allows the user to post tweets and photos that can be examined by other user. So Twitter is chosen as the source for event photo retrieval because of its popularity and data mining. Twitter has unique characteristics that include lot information on different actions in the real world. It is unique from other social media websites. Twitter users can be observed as distributed "social sensors" which report what currently happens over the world. By monitoring the Twitter stream, many of tweets contain not only text messages but also photos. In general, photos can explain what currently happens much more intuitively than tweets. By using such scattered picture sensors effectively; we can know what kind of events happens over the world at this instant visually and intuitively (Kaneko et al. 2015). Previously, the researchers have implemented more work on Twitter event detection and also Twitter have been extensively studied as a sensor distributer of real world trends and events, nearly all of them are based on tweet analysis and their outputs are usually event keywords with their locations and time. But in this research, tweet as well as photo are discussed and analyzed to propose an effective Twitter event photo retrieval method for understanding what happens currently over the world. Here, we identify the events by visual as well as textual information and geo-location information. This is the innovation of our research work.