NATURAL LANGUAGE PROCESSING
Overview
Natural Language Processing (NLP) enables computers to process, understand, utilize and response to human natural languages. Although related research has been developed for many years, it is still quite difficult to reach the human level. The breakthrough of artificial intelligence technology in recent years has brought new opportunities for NLP. TL facilitates the development of innovative applications, which helps understand and extract information from local documents and stimulate creative works cooperated with people. We have also been developing dialog systems with the ability of multi-turn conversation and knowledge question and answering. The NLP and conversational AI technologies we developed are applied to innovative applications such as intelligent customer service, virtual assistants, smart robots, and text analysis/generation, and help digital transformation in many different industrial sectors.NATURAL LANGUAGE PROCESSING
CORE TECHNOLOGY
- Text Understanding and Generation
- Conversational AI
Text Understanding and Generation
Conversational AI
NATURAL LANGUAGE PROCESSING
Application Status
Text Understanding and Generation：Based on the state-of-the-art architecture of pre-trained models, we have developed our models by text understanding on numerous localized corpus to extract information for generation. Our models outperformed that of international manufacturers and was announced at the 2022 NVIDIA GTC Conference. Our models are applied to analyze online public opinion and customer service dialogues, which promptly induce popular topics, summaries and sentiment analysis. Additionally, our models are used in the judicial sentencing factor identification, manuscript generation, tariffs and trade goods classification, and commodity name integration, etc., and have excellent effects on human-aI co-creation.
Conversational AI：We have developed the key technologies of language understanding, multi-turn conversation, dialog state tracking, response generation and chatbots with capabilities of simulating people conversation, so that users can naturally talk and interact with the AI agent in voice or text to seek information or complete specific tasks. Our chatbots are widely adopted in many applications. For example, it serves as virtual assistants in smart speakers, realizes voice shopping, assists with food or table booking, evolves into a switchboard assistant or calls out for payment reminder, etc. We are currently contributing to more sophisticated and professional business conversations and knowledge question and answering applications.