Professor Ji Rongrong shared with the audience the AI research background, research results report, interdisciplinary research and application discussion based on the research problems of different disciplines. First of all, Professor Ji Rongrong briefly introduced the research background related to artificial intelligence. Professor Ji pointed out that the scientific significance of artificial intelligence research lies in the realization of general AI applications through cross-media analysis and calculation of vision, text and voice, and effective content management and knowledge mining for massive Internet multimedia data. In recent years, with the development of technology, an important research trend in industry and academia is to carry out large-scale cross modal pre-training research, using massive multimedia data represented by images and texts to train super large AI models, so as to realize real multi-modal and multi-skill applications.

Next, Professor Ji presented some research results of their research team in the direction of cross media analysis and computing. First, with regard to the task of cross media emotion analysis, Professor Ji Rongrong's team transformed abstract emotions into concrete visual semantic detection by combining psychological emotion concepts with common things. Based on this, with the support of the algorithm, AI can realize real-time monitoring and trend prediction of public opinion events, as well as real-time analysis and detection of social network content in specific areas. In terms of the task of looking at pictures and speaking, Professor Ji's team can not only accurately and comprehensively generate the corresponding natural language description according to the given image, but also accurately erase the image information in the cross modal attack and defense field, so as to avoid sensitive visual information being recognized by AI.

In addition, combined with the research background of management discipline, Professor Ji Rongrong also introduced several related interdisciplinary research topics. Taking the financial sentiment dictionary as an example, because the sentiment of financial texts actually has a very obvious impact on the stock market, but there was no special Chinese sentiment dictionary before, especially for informal text occasions such as social media, post bars, forums, etc., Professor Ji's team used the in-depth learning method to build a financial sentiment dictionary for informal text analysis. At the same time, Professor Ji Rongrong's team also conducted Modeling Research on the innate characteristics of senior executives, objectively evaluated the appearance of senior executives by using the deep learning algorithm, and assisted in the processing of voice features to explore their impact on the IPO market.

Based on the existing technology and demand, Professor Ji discussed the future application of artificial intelligence. Ji believes that at present, cross media analysis and computing plays a major role as a tool for data analysis and knowledge retrieval in the application scenarios of the management discipline. Looking forward to the future, with the deepening of interdisciplinary research, AI may have more intelligent application scenarios, such as intelligent writing and reading of enterprise documents; General AI robot application with professional knowledge and character intention recognition (lying, fraud), etc.

Finally, Professor Ji Rongrong had an exchange and discussion with online and offline teachers and students. The lecture was successfully concluded!