Research On The Utility Of An Advanced ACa In Context-Aware Tour Planning System Research On The Utility Of An Advanced ACA In Context-Aware Tour Planning system Fudan University, Shanghai, 200433, PRO Abstract. Nowadays, tourism has become an irreplaceable part of our daily life. Context-aware service, which acquires contextual information of user and environment, provides high level of customization in mobile commerce, and context- implemented the prototype system following information system design theories of context-aware system, and finale p aware tour guide system brings convenience to travellers around the world. In this paper, we improved the performance of Ant Colony Algorithm(ACA)by focusing on user preferences. Then, in order to research on its utility performance, we validated the improvement by user testing and survey, to demonstrate the significant theoretical and practical value of our research Keywords: Ant Colony Algorithm; Context-aware Service, Tour Planning; Information System Design 1. Introduction he mobile service based on context awareness is a ne lication research area. and the advantage of on and service SS service"[1] in many literatures, and how to provide proper context awareness service to CAIs aviors of ants. mation no ingo the Aca a nss arepabsent tine ohe eutimtiz ed ouit romsprme to om tdestinatiy an important part in solving operation research problems and for much attention of scholars all over the world. They established impressing achievements, and the solution to tour planning problem is continuously updating, providing increasing convenience to people on their trips B. The value of ol Dur research focuses or int Colony a ackground of context awareness service aking context-aware tour as the real situation context. "The idea is to take into account the contex on, customer flow etc, as part hting factors of the ants when searching for the shortest tour, thus improving the performance in customer satisfaction of original ACA in solving the tour planning problem nes in improving the algorithm performance of ACA, our innovation focuses results, in aim of chasing a tighter The improved ACa adopts context information directly into the calculation of the algorithm nstead of simply refer to the data as an outer part of the function. This guaranteed the full use of contextual information dingly with our impi refore, our can be evaluated and validated In this research we bi an advanced aca A with better user customization 2. Literature review A. Ant Colony algorithm Ant Colony Algorithm(ACA) was firstly derived from the ant colony functioning, and the"Ant System"was to be a potential basic algorithm solution to many optimization problems( Colorni al. 1991)3]. The algorithm was then applied to various kinds of combinatorial proble ding the tour planning problem such as Travelling Salesman Problem(TSP), the Vehicle Routing Problem(VRP), etc. [4 In ACA, the ants of a specific region randomly walk among spots to find a route for g food. While walking. they send out pheromone, called"information factor, along the route that can attract all the other ants of the colony to come The earliest use of ACa in tour planning can be described as follows: Let n represents the number of all spots, is the total number nt colony, (ij=1, 2, "",)denotes the d and spot j: let be the t generation of information ween spot i and spot j. At the begin density of information fa r on every route is a constant val that ant k has visited during its walk up till now, and means at til have(k=1, 2, " ,) which is the set of all spots state transit probability for ant k to transit from spot i to spot j. Then we have: 16,71 rule of ants'choosing to transit from one spot to another obeys the fake random rule. moreov ver. whenever an ant chooses its next spot, the algorithm generates a random value within [0, 1], and the system decides its transit direction After time m, the ant finishes one round of travelling among all spots, the system updates the information factors on each route, will gradually decrease by time, and its decreasing rate is defined as. Finally, ants will find the shortest route from home to food according to the density of information factor.Research On The Utility Of An Advanced ACA In Context-Aware Tour Planning System Research On The Utility Of An Advanced ACA In Context-Aware Tour Planning System Fudan University, Shanghai, 200433, PRC Abstract. Nowadays, tourism has become an irreplaceable part of our daily life. Context-aware service, which acquires contextual information of user and environment, provides high level of customization in mobile commerce, and context￾aware tour guide system brings convenience to travellers around the world. In this paper, we improved the performance of Ant Colony Algorithm (ACA) by focusing on user preferences. Then, in order to research on its utility performance, we implemented the prototype system following information system design theories of context-aware system, and finally validated the improvement by user testing and survey, to demonstrate the significant theoretical and practical value of our research. Keywords: Ant Colony Algorithm; Context-aware Service; Tour Planning; Information System Design 1. Introduction A. Research background The mobile service based on context awareness is a new application research area, and the competitive advantage of tourism has gradually changed from resource to the multi-dimensional combination of resource, information and service. The kind of mobile commerce service which is based on context awareness technology is called “context awareness service” [1] in many literatures, and how to provide proper context awareness service to CAIS (context-aware information system) users is a popular research subject. Ant Colony Algorithm (ACA) is a representative of heuristic algorithm inspired from the behaviors of ants. When chasing for food in an open area, ants are able to find the optimized tour from home to the destination. Such method was used in optimization algorithms and has been serving an important part in solving operation research problems and for practical use, including tour-planning problems [3]. The research of improvement and utility of ACA in tour planning has become a hot topic in the field, which attracted much attention of scholars all over the world. They established impressing achievements, and the solution to tour planning problem is continuously updating, providing increasing convenience to people on their trips. B. The value of our research Our research focuses on improvement of the Ant Colony Algorithm under the background of context awareness service, taking context-aware tour planning as the real situation context. The fundamental idea is to take into account the context information, such as weather condition, customer flow etc. , as part of the weighting factors of the ants' information factor when searching for the shortest tour, thus improving the performance in customer satisfaction of original ACA in solving the tour planning problem. 1) Compared with existing researches in improving the algorithm performance of ACA, our innovation focuses on providing better customization rather than higher efficiency or optimization results, in aim of chasing a tighter combination with practical needs and utility value. 2) The improved ACA adopts context information directly into the calculation of the algorithm's core program instead of simply refer to the data as an outer part of the function. This guaranteed the full use of contextual information of our algorithm, as well as a better performance in user customization. 3) Following the existing information system design theory, we designed a set of proper functions accordingly with our improved algorithm to strengthen its advantages in user interaction, user preference analysis, etc. Therefore, our improvement can be evaluated and validated by user testing and survey with specially designed prototype system. In this research, we bring up an advanced ACA with better user customization performance to solve the problem of context-aware tour planning. We hope this paper can offer a trivial but effective advanced ACA, as well as bringing more attention to the study of user experience improvement, to provide a new train of thoughts for the following researches. 2. Literature Review A. Ant Colony Algorithm Ant Colony Algorithm (ACA) was firstly derived from the ant colony functioning, and the "Ant System" was inspired to be a potential basic algorithm solution to many optimization problems (Colorni & al. 1991) [3]. The algorithm was then applied to various kinds of combinatorial problems, including the tour planning problem such as Travelling Salesman Problem (TSP), the Vehicle Routing Problem (VRP), etc. [4] In ACA, the ants of a specific region randomly walk among spots to find a better route for chasing food. While walking, they send out pheromone, called 'information factor', along the route that can attract all the other ants of the colony to come close. The earliest use of ACA in tour planning can be described as follows: Let n represents the number of all spots, is the total number of the ant colony, (i,j = 1,2,…, ) denotes the distance between spot i and spot j；let be the t generation of information factor between spot i and spot j. At the beginning the density of information factor on every route is a constant value . We also have (k = 1,2,…, ), which is the set of all spots that ant k has visited during its walk up till now; and means at time t, the state transit probability for ant k to transit from spot i to spot j. Then we have: [6,7] . (1) The rule of ants’ choosing to transit from one spot to another obeys the fake random rule. Moreover, whenever an ant chooses its next spot, the algorithm generates a random value within [0, 1], and the system decides its transit direction according to the following equation: . (2) After time m, the ant finishes one round of travelling among all spots, the system updates the information factors on each route; will gradually decrease by time, and its decreasing rate is defined as . Finally, ants will find the shortest route from home to food according to the density of information factor