Research And Implementation Of Learning Content Management System And Personalized Recommendation Technology

With the developing of society and the advancing of technology, the application of network has become increasingly extensive in our daily life. Internet technology has become a mainstream technology in today’s society. As an extension and addition of traditional teaching, network teaching has become an important part of teaching in colleges and universities.This thesis, based on the existing research on online learning system from both demotic and foreign institutions and combined with many functions of CBL (Computer Based Learning) network teaching platform, including the existing curriculum design, message collaboration, resource sharing, etc, aims to solve the existing problems in online teaching at present. Using SharePoint content management technology, this thesis has designed and developed INE (Individual Network Education) Learning Content Management System.The system is now ready for many functions including user management, making of multimedia courseware, job production, publishing of learning content, job assignment, access to the teaching content, etc. Learning content creators can not only use the system to develop and design multimedia courseware, combined with text, images, video, audio and other material, but they can also publish the material on CBL platform for online learners.In addition, along with the feature of various types of questions in traditional homework, homework production function of INE makes it convenient for producers to add multimedia material into homework in order to enrich the homework content. Besides, learning tasks are tailored to the needs of different learners. As the CBL platform supports the learning material that meet the SCROM (Sharable Content Object Reference Model) standard, the multimedia courseware content and homework made by INE system all meet the SCROM standard.Along with enrichment of teaching resources, the CBL platform contains a lot of learning materials. In order to achieve the learning goal, learners need to in a short time pick out a most suitable learning path among a sequence of learning objects.This is a big problem which seems impossible to solve for learners. That’s the key problem the INE learning content management system need to solve. The INE can combine teaching resources, provide tools and interactive environment for material producers and learners, it also can satisfy learners’ individualized learning needs,and recommend learning materials based on learners’features. All functions above enable the learners to achieve the best learning effect in the shortest time and greatly improve the service level of online learning.This paper solved the learning path recommendation problem using the ant colony algorithm. The ant colony algorithm enjoys the features of highly parallelism, positive feedback,robustness and cooperativeness,as well as intelligent search and global optimization, therefore, it has been widely used in many subjects, such as combination optimization, communication, integrated electrical route network layout and function optimization, etc. However, several problems need to be settled:the ant colony algorithm is prone to generate prematurity phenomenon; search time is lengthened due to the lack of initial information; selection of algorithm parameter is lack of theoretical basis. This paper has put forward a modified extended ant colony algorithm based on the problem during the application of canonical ant colony algorithm.This paper, by consulting a large number of domestic and foreign references, has proposed the core idea of extended ant colony algorithm--finding the most appropriate learning path for learners. That means choose and recommend learning paths, which match the learners’knowledge and learning style best, and were evaluated as the best, among multiple learning paths constructed by a lot of learning content. Based on learner and learning object’s feature, the extended ant colony algorithm has identified the group similar to learners’ features. With the choice of learning path being guided by similar learners’evaluation on learning object and their matching relation, the positive feedback of valuable information has been strengthened and the algorithm can be quickly converged to the optimal solution at the early stage. As the parameter selection of ant colony algorithm has a significant influence and different parameters are tightly coupled, this paper used simulation experiment to research the relationship between each parameter and performance of extended ant colony algorithm. This paper used genetic algorithm to optimize parameter combination in order to achieve better performance of extended ant colony algorithm. At last, this paper conducted a simulation experiment to verify the feasibility and effectiveness of extended ant colony algorithm. The experiment result showed that the algorithm achieved ideal solution and improved convergence speed. In a word, using extended ant colony algorithm to recommend learning path can satisfy learners’ individualized learning need.