Research Of Adaptive Learning Recommend System Based On The PSGA Algorithm

Currently, the web teaching has great room for development, but also faces enormous challenges-there is a huge contradiction between rich educational resources on web and limited personal ability, therefore, the design and development of adaptive learning system has become one of the hotspots of the web education.Intelligent recommendation is the best way of solving the validity of the information browsing in the current internet learning system. The recommend learning path which can meet the individual needs of learners is a typical combinatorial optimization problem, we use the genetic algorithm which has simple structure and prominent ability of solving practical problem to achieve. Genetic algorithm is easy for early maturity, weak in local search, so this paper proposes an improved genetic algorithm combined with predatory search strategy (PSGA), and implement it on the systematic design and realization of adaptive learning path.The PSGA algorithm proposed in this paper, makes a good balance between the global exploration ability and local development capacity. In the global search, in order to adjust the population diversity and change prematurity, it uses adaptive changing genetic crossover and mutation strategies; in the local search, it explores the predatory search strategy in neighborhood and limits scope, and achieves mature convergence of the algorithm.Based on the research of the PSGA algorithm, the overall recommend process of the adaptive learning path is described thoroughly in this paper, such as the process of chromosome coding, fitness evaluation, genetic operators" selection, predatory local search, then demonstrate the effectiveness of the program through experiments. We develop a adaptive learning system about the C language teaching course, and ultimately reach the goal of recommending a learning path based on the learner's own learning goals, knowledge structure, cognitive level, differences in personal preferences, etc.The problem of recommending a personalized learning path is a simplified model of many practical problems, the framework of PSGA algorithm is flexible and generic, which all can be used in some similar areas. We can choose any kind of genetic algorithm and search strategy according to the specific field to solve the problem.