Convolutional Neural Networks in Human Resource Information Systems for Employee Sentiment Analysis

Emotion management is an important part of enterprise human resource management, and this paper analyzes enterprise employees’ emotions with the help of convolutional neural network technology. The face detection technology is used to detect the coordinates of the key points of the face, and the key point features of the face are extracted based on the one-dimensional convolutional neural network Plain-PD, which is combined with the ResNet network and S3D network to complete the task of recognizing the facial expression. We construct the emotion recognition algorithm using convolutional neural network technology and train the model using the Adam gradient descent method, based on the collected and processed EEG information. Based on emotion recognition, the emotions are classified by the BiLSTM module and multi-head attention mechanism. In all tasks, the emotion analysis model constructed in this paper is superior to other models, with an accuracy of over 90% for both emotion recognition and emotion classification. In the case analysis of Enterprise H, the accuracy of this paper’s model in recognizing employees’ emotions reached 91.36%. At the same time, this paper, with the help of the model of the enterprise employees for emotion recognition and screening, according to the bad emotion screening criteria, shows that there are obvious bad emotions. Approximately 17.47% of employees fall into this category, necessitating the establishment of an early warning system and a focus on emotional management within the enterprise.