Recommendation System Utilizing Multilateral Evaluation Data-based Passport Index with Machine Learning and Its Methods

摘要

The present invention relates to a recommendation system and method based on machine learning technology for measuring total preference according to interactive evaluation data, and more particularly, to data collected when one entity evaluates another entity, and other entities to evaluate themselves. By combining all the data collected at the time, it processes its own preferences and the preferences of other objects with respect to itself with machine learning technology, and provides recommendations that can be interactively satisfied between the recommended object and the recommended object according to the result. It relates to systems and methods. The recommendation system based on machine learning technology for measuring total preference according to the interactive evaluation data of the present invention is a receiving interface (S100) that receives evaluation and target data as input, and a complex data processing interface (S110) that links the input data. , the data processing interface (S110) is composed of a database interface (S120) for inquiring past data to store and normalize data while the interworking is in progress, and a table interface for outputting the result value after the interworking process is completed (S130). The data processing interface (S110) of the system includes a script (S111) for collecting evaluation data, a script (S112) measuring the generation time of the collected evaluation data, a script (S115) for collecting data to be evaluated, generation of the data to be evaluated It is formed of a script (S114) for measuring time, a script (S113) for inquiring evaluation data and past data of an entity to which the data to be evaluated belongs, adjusting it to the value of new data, reflecting and storing it. The present invention maintains the advantages of accuracy and data processing efficiency of the machine learning-based recommendation system widely used in the prior art recommendation system as it is, and at the same time improves the mutual satisfaction, which has not been solved in the existing recommendation system, by interactive evaluation data. It is characterized in that it is possible to achieve mutual satisfaction with the recommendation of the recommended entity and the recommended entity by solving it with a measurement technique. (index word) Interactive evaluation, total preference, mutual satisfaction recommendation system, multi-layer machine learning