Research On Personalized Recommendation Model And Its Application Based On Context-aware

The rapid development of Web2.0 enables Internet content creation and sharing become easier. Every day, there are a lot of pictures, music and videos posted to the Internet. A flood of information provides consumers with rich content, while it becomes difficult for consumers to filter information. Recommended system through the study of consumer interest, personalized calculations and found the point of consumer interest, then guide consumers to find their own information needs. It is a very promising way to solve the current problem of information overload, as an important means about information filtering. Recommended systems in the field of traditional models are often recommended to consider the similarity between consumers and products, and the context in which they consider less, for example, location, time, surrounded people, weather, means of behaviour and so on. The result is that the recommendation cannot be satisfied the consumers. Therefore, an accurate understanding of the consumer context information and put the relevant contextual information to the recommendation model will be key steps to design a recommendation system.Aiming at the current problems existing in electronic commerce products recommended, this paper will integrate contextual information into recommendation process. Then it will by the means of ontology technology, to create personalized product recommendation model based on context-aware. This paper will focus on the construction of situational modeling and ontology, and gives examples of modeling. In this paper, the work accomplished as follows.(1) This paper summarises the based context-aware personalized recommendation of the literature review and theoretical knowledge, concludes the advantages and disadvantages of common methods, and put forward the viewpoint of this paper based on previous studies.(2) This paper analyzes the context factors in e-commerce, and discussed the meaning of situational awareness. Referenced H.Lieberman and others point of view, the situation elements exploded for three parts, which are element of consumers, element of environment, and element of applications.(3) Construction ideas of the situation ontology and domain ontology are analyzed. This paper is based on seven-step method and use LESSM ontology modeling method for modeling situations. A specific example is given to explain the model building process. At the same time, context of multi- granularity partitioning will be given. For building domain ontology, rough set is used to reduce the redundant attribute of domain ontology. It will be easier for users to find real y interested product properties.(4) This paper builds personalized product recommendation model based on context-aware. The basic process model runs are introduced and focuses on key technologies to achieve personalized recommendation. That is synthesis update algorithm and similarity algorithm of the situation.(5) Personalized recommendation model based on context-aware are applied to the tourism e-commerce. Verifed the key attributes of domain ontology after reduction and context ontology similarity matching can improve the recommendation accuracy.