The Context-Aware Crisis Warning Service And Route Recommendation Service For The Visually Impaired Student

China is the worst-hit area of visually impaired patients, has the largest number of patients in the world. The limitation of the vision leads to visually impaired students can’t observe the road when walking, thus visually impaired is one of the most inconvenient crowd in the travel inconvenience. Some patients with blindness due to illness or accident, however, have received a high degree of education, especially some students who is unfortunate to be visually impaired, they hope to be back to campus for mastering some skills to make a living with a positive attitude to face life strongly. Our study aims to help the visually impaired students quickly adapt to the campus environment in order to study and life conveniently.Based on the above analysis, this paper proposes the early crisis warning service of context awareness and route recommendation service for visually impaired students, the main research content includes the following several aspects:Firstly, we take four elements of context as a starting point, and define all contexts which is related to the visually impaired students, then define ontology classes, properties and other related information by constructing the campus context-aware ontology model, we use current mainstream open-source ontology development tool—Protege to construct this ontology model, ontology model of campus is the foundation of context-aware early warning service. For the imprecise and uncertain knowledge of ontology model, we adopt the fuzzy theory. If the fuzzy campus ontology contains fuzzy words or expressions, the degree that element attach to fuzzy subsets will be decided by mood operator:Hλ and membership functions.Secondly, in order to acquire the ability to aware and predict a crisis, this paper builds a critical context found rules, and puts forward two-level semantic reasoning strategy: early warning mechanism based on general inference rules or custom inference rules. Reasoning experiment will use Jess reasoning machine to prove whether the temperature SWRL rules are correct or not. Then we fulfill the reasoning test by the four steps in Jess reasoning engine, the reasoning results meets our expectations.Thirdly, we focus on the personalized requirement of the visually impaired students, and then we put forward a framework of context-aware services for them. The knowledge query layer of the framework reasons on ontology instance and outputs visually impaired students’current conditions. Context application layer pushes the users’ interested routes based on CAMF—DC recommendation model. CAMF—C recommendation model can also realize route recommendations, yet CAMF—DC is better than CAMF—C in visually impaired students’route recommendation, because the learning algorithm of the former is more suitable to the actual scene. In the experimental section, we obtain the real data in two ways:field research and mobile web page, then we use the RMSE and ASP index evaluate recommendation effectiveness of three recommendation models, the experiment result indicates that CAMF—DC possesses better recommendation effectiveness.