| Diet is one of the basic conditions for human survival and development,and it is also a topic often discussed in people’s daily life.With the increasing improvement of China’s economic level and the improvement of people’s living standards,people’s eating habits have begun to change.In order to improve people’s quality of life,it is necessary to have a scientific and reasonable healthy diet.At the same time,healthy diet is not only conducive to maintaining health,but also can prevent and reduce the occurrence of chronic diseases.However,the healthy diet information available on the Internet is growing rapidly,and distributed in multiple sources.It is very timeconsuming for users to understand healthy diet through Internet data.Users need to search data on multiple platforms,select and integrate information.Therefore,there is a need for a healthy diet recommendation system that integrates various data sources in the field of healthy diet to provide people with effective knowledge of healthy diet and healthy diet recommendation.The main research includes:Firstly,the thesis studies the knowledge model of catering recommendation based on ontology.The Ontology Conceptual Model of healthy catering field is constructed by introducing the relevant knowledge of TCM constitution theory.Forming a standardized representation of catering ontology makes it easy to reconstruct and reuse the relevant knowledge of catering,and facilitate the expansion of knowledge and the construction of knowledge map,thus providing a knowledge base for healthy diet recommendation services.Secondly,this thesis studies ontology relation extraction method in text data.Attention mechanism is introduced to solve the ontology relation extraction problem in text data.The performance of the model is analyzed through experiments to illustrate the effectiveness of the model in relation type extraction.Thirdly,this thesis studies the construction of health catering knowledge map driven by multi-source data.Research on knowledge atlas construction of health catering driven by multi-source data,acquiring knowledge from multiple data sources through crawler technology,further expanding knowledge by putting forward entity relationship from text data,realizing data integration of different types,establishing knowledge atlas of health catering field,and forming unified management of knowledge.Finally,this thesis studies the recommendation model of catering O2O.The knowledge map of healthy catering field is used as the basis of reasoning rule base.By establishing semantic reasoning rules and reasoning optimization in healthy catering field,knowledge-based catering recommendation is provided.The recommendation model is applied to the catering O2O platform,and the design and implementation of the catering O2O platform system are put forward.It has a certain practical value for the operators of the catering O2O platform. |
