The project aims to design an intelligent question-answering system for packaging field based on knowledge graph to solve the problems existing in the packaging industry, such as the long industrial chain, large and scattered data, and inaccurate knowledge information retrieval. Techniques such as knowledge graph, intelligent question-answering, natural language processing, deep learning, and personalized recommendation have been adopted for the collection and aggregation, knowledge extraction, and fusion computing on various packaging data on the Internet. A hybrid intelligent question-answering system for packaging field is finally formed. The main functions of the system include knowledge graph, text similarity matching, image recognition and automatic question-answering, which realizes the knowledge card, semantic search and immersive question-answering of big data in the packaging field. With the system, it is handy to obtain the packaging-relevant data on-demand from multi-view and multi-dimension at one-stop in no time through the question-answering mode, thus, enabling the industry to be digitized, information-based, and intelligent.