SmartBullets: A Cloud-Assisted Bullet Screen Filter based on Deep Learning

Bullet-screen is a technique that enables the website users to send real-time comment 'bullet' cross the screen. Since all the comments are shown on the screen publicly and simultaneously, low-quality bullets will reduce the watching enjoyment of the users. In this paper, we present SmartBullets, a user-centered bullet-screen filter based on deep learning techniques. A convolutional neural network is trained as the classifier to remove low-quality bullet comments. Moreover, to increase the scalability of our model, we employ a cloud-assisted framework by developing a backend cloud server and a front-end browser extension. We evaluate our framework with forty volunteers and the result shows that our framework can effectively remove the low-quality bullets and improve the overall watching experience of viewers.