Seite - 182 - in Intelligent Environments 2019 - Workshop Proceedings of the 15th International Conference on Intelligent Environments

Therehavebeenproposed somemethods that consider the trade-offwithother fac- tors. Shiraishi et al. [1] proposed amethodwhich recommends tour routes considering the trade-offbetweensatisfactionand time.Tamashiroetal. [13]definedmulti-objective optimal routingproblems for sightseeingbyextending theoptimal routingproblem and considereda trade-offbetween thevalueof the tourand the requiredmoney.Thesestud- ies, however, consider only a single factor conflicting the satisfaction andmultiple con- flicting factors arenot considered. Users take into account not only satisfaction/experience by the tour but also a bal- ancebetweenitandotherfactorssuchasmoney, timeandstaminaduringthetour.There- fore tourplannersmustconsider/comparemultiplePoIs in termsof the trade-offbetween the satisfactionand the resourceconsumption. Inorder tohelpusersmakedecisionforagoodsightseeingtour,wemodel thisprob- lem as amulti-objective optimizationmethodwithmoney, time and stamina consump- tion of a tourist and satisfaction degree obtained as independent variables. PSO (Parti- cleSwarmOptimization) andMOGA(Multi-ObjectiveGeneticAlgorithm) are popular methodsforsolvingmulti-objectiveoptimizationproblems.SomealgorithmsforMOGA have been proposed so far [14,3]. Among themNSGA-II is one of themost popular algorithmswhichcanderivediverse solutions. The final goal of this work is to presentmultiple diverse tours to users. For this purpose, thediversityofsolutionsisveryimportant.Weconsider threeoffactors:money, time and stamina as resources retained by users, and present diverse solutions to users bysolving the toursearchproblemasamulti-objectiveoptimizationproblem taking into account trade-offbetween these resourcesandsatisfaction. 3. DiverseTourRoutesSearchProblem 3.1. Problem Weassume that user resources consumed by sightseeing activities andmovements be- tweenPoIs are (1) stamina, (2) time and (3)money.Moreover, to simplify theproblem, the tourist’s purpose of the sightseeing tour is only obtaining satisfaction by the sight- seeing. Our target problem is amulti-objective optimization problem to derive tour plans that have higher satisfaction with smaller consumption of resources consisting of stamina, timeandmoney. Thevaluesof fourvariablescorresponding tostamina, time,moneyandsatisfaction varydependingon the initial valuesassigned to thesevariables and increase/decreaseof themat eachPoIandeachmove included in the tourplanunderconsideration. 3.2. ProblemFormulation LetX={x0,...,xn} denote the set of tourist spots (PoIs). Let rn=(mn,tn,sn)T denote thevector of thevaluesof the remaining resources after enjoying sightseeingat eachof n spots. Here,mi, ti and si denote the remaining amounts ofmoney, time and stamina, respectivelyafterenjoying i-thvisitingspot inX.Letr0=(m0,t0,s0)T denote thevector of initial valuesassigned to resourcevariables,wherem0, t0 and s0 are initial valuesand Y.Hiranoetal. /AMethod forGeneratingMultipleTourRoutes182