Event Detection and Classification Using Deep Compressed Convolutional Neural Network

Recently, the number of different kinds events on social media platforms show a tremendous increase in each second. Hence, event detection holds very important role current scenario. However, is challenging information technology (IT). Several machine learning-based approaches are established for process, but it generates high error and makes various loss, affecting system’s performance. Thus, proposed work introduces new strategy based deep learning architecture. In this, both text image data utilized detection. The procedures databases pre-processing, extraction classification. pre-processed using four methods: lower case filter, tokenization, stemming, stop word filter. An adaptive median filter (AMF) pre-processing data. After stage, feature performed image-based which most useful features extracted. Finally, varied detected classified Deep Compressed Convolutional Neural Network (DCCNN). entire implemented PYTHON platform. efficiency model measured by evaluating performance metrics such as accuracy, recall, precision F-measure. simulation validation exhibits that classification method attains an improved accuracy 97.1%, obtained about 95.06%, recall value 91.69%, f-measure 93.35%. efficacy proved comparing attained results with state-of-the-art techniques.