Until now, researches have brought up many improvements on ACa in so problem. These includes the Max-Min Ant System Algorithm [8], or using time-varied function Q(t) to replace ation factor density p 9 and also the definition of intelligent ants' in literature better overall optimizing ability [71 However, improvements for utility purpose and better customer experience are still in an early stage without consens ethods. One of the key problems that is puzzling researchers of context-aware tour planning information system design i how to find a best way of utilizing context information in the functioning design, in order to provide best user experience oa to the core computing part of the advanced aCa to get better customized results B. Context-awareness service used to ch Existing researches of context-aware systems are primarily confined to the definition and acquirement of contextual information, or bring forth the structure of context awareness services, experimenting on prototype systems or ning only one small pect of context-aware service or confined to one specific simple application. Besides, when combined with tou context awareness service can give birth to a new field of research. Tour guide system is one of the earliest applications of Thisomype systems abstract context information only ine context and directly present to the user in their designed functions. 3. Advanced ACA and function design A. Our improvement on ACA utilization generally speaking, existing researches on ACA, context awareness services and tou of reference for later researches. However, most of the improvement of ACA me results, but few was done on finding a more systems. We adopt context information into the calculation of the algorit core program in our ad ACa instead of simply refer to the data as an outer part of the function context information value into the calculation and optimizing process In our advance factors, but no one has directly pu Moreover, previous researches into ACA had done adjustments on the information weather and customer flow, are given corresponding values and standardized to be a or of the information system sent out by the ants. This guaranteed owing the existing information system design theory, we designed a set of proper functions accordingly with ou improved algorithm to strengthen its advantages in user interaction, user preference analysis, etc. in order to evaluate and validate the algorithm utility by user testing and survey on specially designed prototype system B. Advanced ACA In tour route planning problem, how to make full use of the context information is of crucial importance to the customization performance of the system. In order to emphasize the influence of context i results, we quantized the context information such as weather, customer flow, gave them a standardized value within and multiplied as an weighing factor in the ant colony's information factor The pseudo code of the advanced ant colony algorithm is as follows: ce of the sites ec spots[l): //ask the user to set key points if( tmp>N_CITY_COUNT II tmp < 1)&& if(Available) Recommendspoto//system recommending process for (in spots[a]= rec spots]: l/ recommendation complete //Main function nitializeCities. For(int i>k: k<=n: i++)//for all the sites For(int j=l: j<=m; j ++)//the number of ants is m xtinfodbTEMP) //weighting of weather and customer flow fresh(: //update the OutputPath(p); //output the optimized route //Multi-choice ou For(bset tour spots alc( cost(p), customerflow(p), recommend(p) Output(ArrayIcost, p) Array[rush, pL, Array[prep, p: Check(repeat); //whether to choose the spots again In the improved algorithm, we weighed the information factor by context information as follows C. User satisfaction function On valuing the user satisfaction on his experience with the advanced ACA prototype system, we referred to sors90004) research results from existing [11] and Huo and miller(2007)[ 12] thought the satisfaction came from spots, service and experience, which wereUntil now, researches have brought up many improvements on ACA in solving tour planning problem. These includes the Max-Min Ant System Algorithm [8], or using time-varied function Q(t) to replace the information factor density p [9], and also the definition of ‘intelligent ants’ in literature [10]. All these improvements focus on higher computing speed or better overall optimizing ability [7]. However, improvements for utility purpose and better customer experience are still in an early stage without consensus methods. One of the key problems that is puzzling researchers of context-aware tour planning information system design is how to find a best way of utilizing context information in the functioning design, in order to provide best user experience in customization and recommendation. Therefore, in this paper we come up with a new way of combining context information into the core computing part of the advanced ACA to get better customized results. B. Context-awareness service Anind K.Dey and Gregory D. Abowd [2] gave a formal definition of “context”: Context is any information that can be used to characterize the situation of an entity. Any entity is a person, place, or object that is considered relevant to the interaction between a user and an application, including the user and applications themselves. Existing researches of context-aware systems are primarily confined to the definition and acquirement of contextual information, or bring forth the structure of context awareness services, experimenting on prototype systems or comparatively simple context-aware service [1,5]. Obviously, these researches are dispersive, concerning only one small aspect of context-aware service or confined to one specific simple application. Besides, when combined with tourism, context awareness service can give birth to a new field of research. Tour guide system is one of the earliest applications of context-aware service. However, these practices only involved simple contextual information, and almost all the recent prototype systems abstract context information from the context and directly present to the user in their designed functions. This may not offer the best effect in user experience and customization. 3. Advanced ACA and function design A. Our improvement on ACA utilization Generally speaking, existing researches on ACA, context awareness services and tour route planning have provided fundamental value of reference for later researches. However, most of the improvement of ACA mentioned above focused on higher efficiency or better optimization results, but few was done on finding a more feasible way of ACA utilization in real functioning systems. We adopt context information into the calculation of the algorithm's core program in our advanced ACA instead of simply refer to the data as an outer part of the function. Moreover, previous researches into ACA had done adjustments on the information factors, but no one has directly put context information value into the calculation and optimizing process. In our advanced ACA, context information, such as weather and customer flow, are given corresponding values and standardized to be a weighing factor of the information system sent out by the ants. This guaranteed the full use of contextual information of our algorithm, as well as a better performance in user customization. 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. in order to evaluate and validate the algorithm utility by user testing and survey on specially designed prototype system. B. Advanced ACA In tour route planning problem, how to make full use of the context information is of crucial importance to the customization performance of the system. In order to emphasize the influence of context information to the optimizing results, we quantized the context information such as weather, customer flow, gave them a standardized value within [1, 2], and multiplied as an weighing factor in the ant colony’s information factor updating process. The pseudo code of the advanced ant colony algorithm is as follows: GetUserInfo(); //get user information GetSpotsInfo(); //get user preference of the sites { Input UserKeySpots(Array_rec_spots[j]); //ask the user to set key points Check if(tmp > N_CITY_COUNT || tmp < 1) && if (Available); GetTotalNum(int); } RecommendSpot() //system recommending process { GetUserInfo(); for (int a = 0; a < total_spot_num; a++) { spots[a] = rec_spots[a]; }// recommendation complete } //Main function InitializeCities; For(int i>k；i<=n；i++) //for all the sites { For(int j=1; j<=m; j++) //the number of ants is m Calc(prob(i,j)) } GetOptimizedPath(p); ContextInfo(dbTEMP); //weighting of weather and customer flow Refresh(); //update the information factor OutputPath(p); //Output the optimized route //Multi-choice output For(bset tour spots) { Calc( cost(p), customerflow(p), recommend(p)); Output(Array[cost,p],Array[rush,p],Array[prep,p]); } Check(repeat); //whether to choose the spots again In the improved algorithm, we weighed the information factor by context information as follows: dbTemp = pow(dbTemp,0.5) * weather[final_spots[j]] * cust_flow_index[final_spots[j]] ; This demonstrates the adjustment of adding an weighing factor to the Ant System. This may not offer higher efficiency or optimizing ability to the algorithm, but for the purpose of reaching an advanced customized context-aware service level and providing better user experiences. C. User satisfaction function On valuing the user satisfaction on his experience with the advanced ACA prototype system, we referred to some research results from existing literatures which focused on motivation, satisfaction and destination loyalty. Harrison (2004) [11] and Huo and Miller (2007) [12] thought the satisfaction came from spots, service and experience, which were