Personalized advertisement push method based on semantic similarity and data mining
Mengjiao Yin · 2021 Third International Conference on Inventive Research in Computing Applications (ICIRCA) · 2021
Personalized advertisement push method based on semantic similarity and data mining is studied in this paper. According to existing research, it is difficult to use one of them alone to make the feature comparison between images sufficient and also comprehensive. In addition, only performing feature extraction and fusion on the global and local scales can hardly break through the bottleneck in retrieval accuracy. Therefore, this research work utilizes the calculation model to extract the relevant attribute information of the product from the forwarded information, and decide whether to include it in the system's recommendation database or exclude it from the sentiment attitude. The designed model is simulated on the existing data sets to verify the general performance. It is reported that the pushing performance is much well improved.