الفهرس 
الغلافOnly 14 pages are availabe for public view
 |  | from | 33 |  |  |
| | | from | 33 | | |
Abstract
Traffic jams and congestions have bad effects on drivers and the whole community. In 2010 alone, the United States of America amassed a total loss of $100.9 billion as a direct result of traffic jam [1]. This figure covers the direct monetary losses associated with the loss of 4.82 billion working hours as well as the burning of 1.94 billion gallons of fuel in during traffic jams.
Traditional solutions for traffic network congestion usually involve constructing new roads or expanding existing ones. The drawbacks of similar approaches include high money cost and residents inconvenience. One approach to handling this problem without suffering from the drawbacks of traditional solution is to use Route Planning systems. Route planning systems aim to improve the user’s experience by advising them on the routes that would lead them to their destinations as fast as possible. But those systems don’t solve the problems of traffic jam and road congestions; instead, they merely work around those occurrences by factoring them in the travel time calculations. As a result, only the system’s registered users benefit from it and even then they don’t get to avoid congestions completely.
Most - if not all - route planning systems focus on finding the optimal/fastest route for its users. Users might be advised to pass through a congested road section if it will take less time to traverse than its alter~atives. This approach ignores the negative effects of congestion on both of the individual and the whole community. Instead of focusing on giving each user the fastest route to his destination, this papers aims to develop a model that would be optimal to the whole community.
The main objective of the proposed model is to improve the efficiency of transportation and to mitigate the damages associated with congestion by giving all registered vehicles simultaneous route suggestions that would avoid congested roads, whether the congestion has already taken place at the time of planning (traffic jam detection) or is expected to occur by the time the vehicle reaches that segment on its trip (traffic jam prediction).
Stanford University’s open source ontology editor, Protege [2] was used to design the ontological model tl7eo save it io OWL format Java programmiog language [3] was used to write the central server’s back end. Apache Jena [4] framework was utilized to allow the Java-based back end to interact with the OWL files and take advantage of other ontology features. Java programming language was also used to write the simulation and RRRS steps that don’t involve ontology-based inference. For Ontology based inference, Jena’s default RDFS reasoned was used. (RRRS is described in more detail in section 4.7) According to the simulation results, the proposed model succeeded in delaying congestion formation by an average of 63.5 minutes. It also succeeded in keeping the roads congestion free for 59.42% more than comparable systems.