Video tagging, paraphrasing, sentiment analysis, and other annotations may help news display, information retrieval, and information clustering, and improve user experience Multilingual and multi-scenario speech, image, video, and text data collection and annotation can be used in chatbot, speech search, cross-modal retrieval, online Q&A, and other intelligent applications