Bill Gates once said, “Language understanding is the crown jewel of artificial intelligence. Advances in natural language processing (NLP, Natural Language Processing) will drive the overall progress of artificial intelligence.”
Time goes by to the 2019 World Artificial Intelligence Conference (WAIC), what level has NLP reached? Replacing journalists writing articles, replacing poets writing poems, and even replacing employees with repetitive tasks. At the NLP theme forum hosted by Daguan Data and Leyan Technology, a number of global NLP experts discussed the problems and future of voice intelligence changing life.
Since the technology of deep neural network was proposed by Professor Hinton in 2006, after more than ten years of development, it has become more and more mature, and the automatic processing of text is facing a very good opportunity and began to subvert various industries.
NLP has given robots the ability to read and think, and robots are no longer the plot of the movie to replace employees with repetitive tasks. According to Chen Yunwen, the founder of Daguan Data, at present, the robot employees launched by Daguan Data have gradually carried out various tasks in scenarios such as business case report generation, smart government administrative approval, financial document inspection and filling.
Customer service is labor-intensive and knowledge-intensive. AI has gradually penetrated into this industry, and now AI customer service has been maturely applied. Leyan Technology has developed the first anthropomorphic AI customer service robot “Leyu Helping People”, which has the ability to assist human customer service in consultation reception, business processing, intelligent recommendation, etc., and has achieved similar Turing test results in this vertical field.
During WAIC this year, Daguan Data launched an AI news assistant, which integrates natural language processing (NLP) and optical character recognition (OCR) technologies, allowing reporters to experience artificial intelligence in the collection and transcription of article materials, automatic summary writing, and content classification. It is convenient to help reporters quickly output news reports.
The “IT Times” reporter found that the assistant can quickly make five-character and seven-character quatrains that conform to the norm. The quatrains are flexible and light, and the tones are relatively flat. According to Chen Yunwen, the news assistant has conducted “corpus” learning on more than 50,000 Tang poems, and through the generation of semantic analysis models and continuous training, they finally have the ability to “export into chapters”.
Leyan Technology also launched the AI intelligent tax consulting system “Xiaohui Xiaohui” at this World Artificial Intelligence Conference. The system is a typical application of Leyan Technology’s cross-language cognitive computing platform empowered in the field of government affairs, and can provide consulting services to every taxpayer.
Language is an important way of human communication, how will artificial intelligence decipher the code of human language? NLP has achieved a qualitative leap. Chen Yunwen said that the ability of robots to process repetitive and mechanical words will soon surpass that of humans. Ten years from now, more than 50% of basic office work will be done by text flow automation robots.
Not long ago, Daguan Data released the first domestically developed Daguan Intelligent RPA that integrates NLP and OCR. Robotic workers are born by letting computers learn and mimic the steps humans take to process tasks and assist with repetitive clerical tasks.
Linguistic intelligence, at the heart of the development of artificial intelligence, will change the way we read, review and produce document content at work. With the development and application of natural language processing technology, there is a lot of room for imagination in the future of human-machine collaboration.