Movie Recommendation Algorithm Based on Sentiment Analysis and Topic Model

Traditional recommendation algorithms often face challenges such as data sparsity and a lack of diversity in recommendation results. To address these issues, this paper proposes a novel recommendation algorithm that integrates graph neural networks (GNN) and topic modeling. Initially, a topic model is employed to determine the topic distribution of sentences in reviews, connecting sentences with the same topic. Additionally, a gating unit is introduced to regulate the information flow between the layers of the GNN. To capture the internal and external causes of sentiment changes more effectively, differences in character information are considered in the edge types. This approach enriches both the user interest model and the product feature model based on sentiment density. In the experimental performance on the Douban movie review dataset, the precision, recall and F-measure of the new algorithm are 91.25%, 89.27% and 89.74%, respectively.