원문정보
초록
영어
With the rapid growth of personal website influence, the advertisement placing has become an important investment in personal websites. But in order to accurate the advertisement placing, the specific quest for the specific users with their specific interesting need to be concerned. Acquiring, preprocessing and classifying consumption intention of the released information that published in the personal websites is the main task of this essay. We regard consumption intention mining as a binary classification problem, and extract multi-dimensional features from the raw corpus. Finally, we propose models based on SVM, Naïve Bayes and deep learning to solve the consumption intention classification problem. The experimental result shows that the deep learning based method achieves the highest F-measure.
목차
1. Introduction
2. Intent Analysis and SVM Classification Theory
3. Classification model for Intent Mining based on SVM
3.1 The Fusion of Multiple Attribute based on SVM
3.2 Classification Model of Intent Mining based on Fusion Multiple Attributes based on SVM
4. Experiment
5. Conclusion
Acknowledgement
Reference