Real-Time Hashtag based Event Detection Model with Sentiment Analysis for Recommending user Tweets

Recently, the usage of Social Networking Sites (SNS) has increased tremendously. People use online social networking applications like Twitter to specify their opinions and feelings about many topics which they produce an awesome amount of data every day. This generates an unlimited amount of data requiring significant effort to read tweets relevant to the user preference. In this paper, we proposed a model “Real Time Hash tag based Event Detection for Tweet Recommendation” on the level of individual posts. This model recommends Events at various levels from the foremost abstracted to the foremost definite. Firstly, pre-processing techniques like stop word removal, frequent word removal, stemming, lemmatization, lower-case conversation are applied. The proposed HashTag Event Detection (HTED) algorithm is applied to detect events corresponding to the tweets and find the sentiment polarity of each user interested in a particular event. This represents indirectly the interest of the particular user. Finally, the Support vector decomposition (SVD) technique is applied to predict the user preference of events in social media streams using hashtags in posts collected from Twitter. Experiments are performed using our dataset extracted over 1-Lakh tweets of different users. The performance of the HTED system is evaluated in terms of the Receiver Operating Systems curve and mean average recall curve. It is observed that the proposed HTED model performs better than the traditional Random Recommendation and Popularity Based Recommendation for users using hashtags.