| With the development of network technology,online education has become an important and popular way of learning.Thus people have higher requirement on the network learning environment.Learners have been unsatisfied with the traditional retrieval methods for obtaining the learning resource objects.At the same time,the rapid development of web2.0 made the sharing of learning resource objects overload seriously.It is difficult for learners to find suitable learning resource objects in the massive amount of the learning resou,rce objects.In order to solve these problems,many experts and scholars are devoted to the research of personalized recommendation.At present,collaborative filtering is popular in the field of the electronic commerce and it has gotten mature theoretical and technical support.But the research on the personalized recommendation of learning resources is still in the early stage at home and abroad,.Through a large number of literature analyses,we not only found that different learners will choose different learning resourse object because of different knowledge background but found that the recommendation of learning resources and commodity is also different.Study is a progressive process.So there exists a certain associated relationship between the learning objects.This kind of relationship should be set up by itself autonomously and intelligently rather than relying on the large number of manual annotations.In this paper,we cluster the learners with similar feature by using the idea of collaborative filtering firstly.The learners in the same cluster are thought that they have similar learing background and interest orientation and on the contrary,the learing background and interest orientation of the learners in the different cluster will be different.By this way,we can classify the learners according to their different learning background.Secondly,based on ant colony algorithm,we know that ants can find the shortest path between the food and the nest no matter how far the food is.And in the real life,the trail of learners' learning activity of likes the path of looking for food of the ants.Thus on the basis of clustering,the ant colony algorithm is applied to the cluster.By this way,we can find out the N learning resource objects with the maximum probability selected by the learners and recommend them to other learners.Finally,the author's proposed algorithm is experimentally verified.The main contents of this paper include the following aspects:(1)In the process of clustering the learners who have the similar feature,the matrix model of the relationship between the learner and the learning resource object tags is established.The matrix model aims at reducing the memory occupancy rate and time complexity.k-means algorithm is used during the clustering process.Because the k-means algorithm is sensitive to the selection of the initial parameters,the algorithm has to be optimized in this paper.(2)Ant colony algorithm is respectively applied on each similar learner clusters.In the process of looking for learning resource objects,an ant is seen as a learner and the food is seen as a learning resource object and the pheromone(a volatile chemical substance left by the ant in the search for food and the chemical substance can affect the path selection of the other ants)is seen as the score which the learner gives to the learning resource object.Then calculations such as the pheromone,the expected transfer probability,the heuristic function design,the update of the pheromone and so on will be adjusted and updated.(3)In order to verify the validity and feasibility of the proposed algorithm,we used the mark data of Douban books.Finally compare the quality of ours recommendation algorithm,the collaborative filtering remmendation and the ant colony algorithm. |
PDF Full Text Request