Research And Implementation On Online Illegal Advertising Identifying Technology

Along with the progress of science and the development of society, electronic commerce attracts more and more advertising operators and consumers. It plays an irreplaceable role in the fields of economy and society and brings great convenience for tens of thousands of people. At the same time, a lot of problems have appeared:the infringement of trademark, the false propaganda, the protection of consumer rights and so on. How to effectively supervise and recognize the illegal internet ads has very important significance.This paper focus on the AIC (administration for industry and commerce) supervision field and research and implement the key technology of the web illegal text advertising intelligent identification. We use the improved text classification algorithm to classify text ads. The semantic features are mined from Wikipedia and then added to the document. Then a new document similarity calculation method is proposed. Clustering process is proceeded to extend the labeled samples.For the illegal words form ads, we propose the improved keyword matching technology and combine the context and named entity to identify the illegal keywords. For the illegal describe sentences form ads, we proposed the illegal advertisement recognition model base on probabilistic potential semantic analysis, considering the short text and the lack of semantic with advertising text. Many experiment results show that the proposed algorithm can improve the effect of the illegal advertisement recognition.Finally, we design and realize the illegal advertisement intelligent inspection system. We introduce the training process of the illegal advertisement recognition model, the data acquisition system and task management and illegal report management platform.