WITHDRAWN: Emotion recognition simulation of Japanese text based on FPGA and neural network

Abstract

Abstract Text emotion recognition is an important task to identify the human reaction, with several solutions in various areas such as data mining, online learning, information filtering systems, Human-Computer Interaction, and psychology. Specific text emotional perception is the most solved problem in the literature. Hidden emotion recognition is to solve this problem, as it is the most difficult problem hidden in ordinary text. Therefore, the solution needs to understand the text in faster action, so the Neural Network (NN) is implemented in this system. The text recognition of Japanese language is difficult in existing methods like processing time is low, accuracy is high and power consumption is low in (Field Programmable Gate Array) FPGA based neural network classifier. This type of Neural Network (NN) analysis is beneficial to Japanese emotions as it implements the classification result so the accuracy is improved. To test this theory, an automatic emotional word recognition system uses the learning and text features and is based on classification. To analyze the performance of the proposed Neural Network (NN) like a flip-flop, power, latency, precision, recall, and F-measure. It turns out that recognition based on classification can significantly increase the result. Performance highlights of these methods and the impact of simple neural network tasks such as reduce the issues this also works for the best performance.