Infrastructure Engineering - Theses
Permanent URI for this collection
Search Results
1 - 2 of 2
-
ItemAgent behavior in peer-to-peer shared ride systems( 2007-01)Shared ride systems match the travel demand of transport a client with the supply of vehicles, or hosts, so that the client find rides to their destinations. A peer-to-peer shared ride system allows a client to find rides in an ad-hoc manner, by negotiating directly with nearby hosts via radio-based communication. Such a peer-to-peer shared ride system has to deal with various types of hosts, such as private cars, taxicabs and mass transit vehicles. Agents, i.e. a client and hosts, have diverse behaviors in such systems. Their different behaviors affect the negotiation process, and consequently the travel choices. Preliminary research (Winter et al. 2005) has investigated peer-to-peer shared ride systems with homogeneous hosts and immobile client. This thesis extends their work to multiple types of agents. It focuses on what are typical agent behaviors in peer-to-peer shared ride systems, and how these behaviors affect negotiation processes in a dynamic transport environment. (For complete abstract open document)
-
ItemInterpreting destination descriptions for navigation services( 2011)People move from one location to another to complete their daily activities. Finding the destination and determining a way that leads to the destination is a demanding task, particularly in an unfamiliar environment. Destination descriptions are referring expressions in natural language that uniquely describe the destination of a route in an environment. As a wayfinder, you may be able to identify your current location, and have some information about the destination, but you may only have coarse spatial knowledge about how to get there. There are local people in the streets, who can understand where you are headed and help you find the way to your destination. However can you still reach your destination efficiently if only asking help from a navigation service? This research intends to answer this question by exploring intelligent navigation services that simulates human behavior and understands requests in natural language. Route communication includes three phases between wayfinders and informants: the initial phase is when wayfinders ask information for directions from the start point to the destination, the center phase where informants offer information, and the last phase is confirmation and closing. While much research in this area has converged on the center phase, the focus of this research is in completing the initial phase. The initial phase of route communication is vital to complete route communication. This is because during the initial phase, a wayfinder and an informant make an agreement on the start point and destination. If an agreement cannot be successfully made, the informant cannot provide directional information and the wayfinder cannot close the communication and act. However, wayfinders often find difficulties with current navigation systems at the initial phase because they cannot deal with destination descriptions in general. This lack of spatial semantics in the fundamental mechanism of navigation systems motivates this research. An intelligent navigation system should have capabilities to imitate human route communication behaviors and make sense of place descriptions at the initial phase. This research firstly explores how humans perceive and express urban environment in natural language, and investigates the structure of destination descriptions. It secondly investigates the solutions to semantically matching destination descriptions with geospatial data. This research then formalizes an interpretation model, which works in the initial phase of route communication with navigation services. The input of the model is a destination description in natural language, i.e., plain English. This model filters the available data using spatial reasoning, and offers the most relevant reference matching the destination description as a result. Collected from Victorian gazetteer data, real data are prepared with all the required attributes for use. A simple natural language processing tool is employed to parse and identify place names and spatial relation entities for the model. Running the automatic interpretation model behind, two experiments are designed to evaluate its performance. The assessment is done by comparing the interpretation of results from human destination descriptions to a state-of-the-art navigation service. It was observed from the results that compared to current interpretation models, a spatially enabled semantic interpretation model of destination description can improve the performance of inferring destinations. This research contributes to the field of spatial cognition in general and to research on spatial reasoning of destination descriptions in particular.