Core Technology


PFG in Chinese semantic recognition, self on large-scale corpus, through the corpus of words found that combine industry vocabulary, and integrate them into a specialized vocabulary, network rich vocabulary of ontology, also joined the large number of car industry, specific rules for the network, very applicable to automobile industry of Chinese semantic analysis. PFG every week to collect tens of thousands of Web corpus, and constantly expand the scale of the body, so that it is closer to the actual application of the scene. In the model, PFG uses the CRF model of semi supervised learning to train the corpus, and obtains the semantic analysis model of the automobile industry. In addition, PFG also improved the segmentation, developed new word detection technology, and build the automotive industry based emotion analysis technology, automatic identification consumers on Car object criticism and review the emotional tendencies, and can automatically on the hot topic of discussion to realize clustering. |
![]() |