The multi-day intermodal travel planning problem (MITPP) is an optimization problem (OP) and it generates the optimal sequences of point-of-interests (POIs) and hotels while searching for the most suitable transport modes between POIs and hotels. Conventional methods and solvers using von Neumann computers provide good approximate solutions to the OPs, but the computation time grows exponentially dealing with large-scale or complex OPs. Meanwhile, Ising machines or quantum annealing machines are non-von Neumann computers that are designed to solve complex OPs. In this paper, we focus on solving the MITPP by a two-phase Ising-based method. The first POI clustering phase aims at generating POIs clusters for sightseeing days and the second POI routing phase generates travel routes for each day with the optimal transport modes. Practical factors such as POI satisfaction, POI duration, hotel fee, and transportation fee are included in the MITPP. We map these elements onto quadratic unconstrained binary optimization (QUBO) models. For evaluation, we use a real-world dataset in Sapporo, Japan. Empirical results confirm that the proposed method can effectively solve the MITPP both in terms of solution quality and execution time and outperforms a conventional solver, a conventional method, and the latest Ising-based method.

Vehicle routing applications are ubiquitous in the field of pick-up and delivery service. We focus on the vehicle routing problem with balanced pick-up called VRPBP which originates from the package pick-up service. The aim of the problem is not only to efficiently explore the shortest travel route but also to balance loads between depots and vehicles. These problems can be regarded as optimization problems, and recent developments in Ising machines, including quantum annealing machines, bring us a new opportunity to solve complex real-world optimization problems. In this paper, a two-phase method and a three-phase method using Ising machines are proposed for solving the VRPBP. As the applicability of current Ising machines is limited due to the small size of Ising spins and connectivities, we partition the complex problem into two or three sub-problems, and the key elements of each sub-problem are mapped onto quadratic unconstrained binary optimization (QUBO) models to fit in the structure of the Ising machines. We first compared the performances of the Ising machine on the standard TSP and CVRP datasets with a conventional state-of-the-art solver and three conventional methods. Then, we evaluated the performances of the proposed methods compared with five conventional method for solving the VRPBP. The results confirm the effectiveness of the two proposed methods in solving vehicle-routing-related optimization problems.

The trip planning problem (TPP) can be formulated as a combinatorial optimization problem that searches for the best route to visit a series of landmarks and hotels. Meanwhile, Ising machines have attracted attention due to their efficiency in solving combinatorial optimization problems. The Ising machines solve the combinatorial optimization problems by transforming the problems into quadratic unconstrained binary optimization (QUBO) models. However, the possible input QUBO size of current Ising machines is quite limited. Thus, it is hard to directly embed a large-scale TPP onto the current Ising machines. In this paper, we propose a novel subQUBO annealing method based on the combined variable selection method to solve the TPP. The proposed method finds a quasi-optimal solution to a large problem by repeatedly partitioning the original QUBO model into small subQUBOs that can be embedded onto the Ising machine. Specifically, to construct a subQUBO, we select variables from the original QUBO model, which have small deviation values. Further, we select variables randomly from the original QUBO model, so as not to fall into the local optimum. We have conducted an evaluation experiment using Ising machines on TPP and confirmed that the proposed method outperforms the state-of-the-art methods in terms of POI satisfaction and POI cost.

Optimum or quasi-optimum solutions of combinatorial optimization problems can be efficiently found by using Ising machines. The graph coloring problem (GCP) is known as a difficult combinatorial problem. Given a graph, each vertex should be assigned a color and any two vertices connected by an edge must not be colored the same. Although methods to map the GCP onto the Ising model are proposed, none of them considers minimizing the number of colors. In this paper, we propose a mapping method of the GCP including minimizing the number of colors and additional constraints to the QUBO (Quadratic Unconstrained Binary Optimization) model. As well as the constraint terms for the GCP, we firstly propose an objective function term that can minimize the number of colors so that the number of used spins cannot increase exponentially. Secondly, we propose two more additional terms so that our proposed method can be applicable to practical consumer applications. The experimental evaluations on an Ising machine showed that the proposed method reduces the number of used colors by up to 75.1% on average compared with the existing baseline method. Considering the additional constraints, the proposed method can effectively find feasible solutions satisfying all the constraints.

An earnings announcement report (EAR) contains the latest information about a company's financial situation and operating performance. Short-term stock price reacts strongly to such information. In this paper, to gain investment return from the short-term price reaction to EARs, we use 28 important variables from EARs and propose an ML-based trading strategy (MLTS) with random forest (RF). Results show that our strategy achieves the highest final investment return of 178.1%.

Trip planning services have been developed along with tourism promotion and information technology evolutions, where we must construct trip routes that simultaneously optimize multi-objective functions such as trip expenses and user satisfaction. Moreover, utilization of past-trip records is essential, because similarities to past-trip records well reflect users' general preferences and tendencies during trip planning. In this paper, we propose a multi-objective trip planning method using ant colony optimization (ACO). By effectively using the pheromones in ACO, we can construct trip routes similar to trip records stored before and the constructed route can reflect users' general preferences. In addition, we vary ants' behaviors in ACO corresponding to various objective functions and hence we can obtain multi-objective trip routes naturally. Experimental results demonstrated that our method outperforms the baseline methods in terms of point-of-interest (POI) satisfaction, POI cost, and past-trip similarity. We also conducted a user study, which clearly indicates that our method obtains high scores through various user questionnaires.

Vehicle routing problems (VRPs) can be solved as optimization problems. Practical applications of the VRPs are involved in various areas including manufacturing, supply chain, and tourism. Conventional approaches using von Neumann computers obtain good approximate solutions to the optimization problems, but conventional approaches show disadvantages of computation costs in large-scale or complex problems due to the combinatorial explosion. Oppositely, Ising machines or quantum annealing machines are non-von Neumann computers that are designed to solve complex optimization problems. In this paper, we propose an Ising-machine based approach for the vehicle routing problem with balanced pick-up (VRPBP). The development of the VRPBP is motivated by postal items pick-up services in the real-world. Our approach includes various features of VRP variants. We propose a 2-phase approach to solve the VRPBP and key elements in each phase are mapped onto quadratic unconstrained binary optimization (QUBO) forms. Specifically, the first phase belongs to the clustering phase which is an extension to the knapsack problem with additional distance and load balancing concerns. The second phase is mapped to the traveling salesman problem. Experimental results of our approach are evaluated in terms of solution quality and computation time compared with conventional approaches.

The multi-day travel planning assists users with realistic travel itineraries by searching for the optimal travel routes through a set of candidate hotels and point-of-interests (POIs). The multi-day travel planning problem (MTPP) can be solved as an optimization problem. Although conventional methods using von Neumann computers obtain good approximate solutions to the optimization problems, large computation costs are required to solve large-scale or complex problems due to the combinatorial explosion. On the other hand, Ising machines or quantum annealing machines are non-von Neumann computers, and those machines are developed to deal with complex optimization problems. In this paper, we propose an Ising-machine-based method for the MTPP. Practical factors of the MTPP include the POI satisfaction, travel expenses, and time limits. Those factors are mapped onto quadratic unconstrained binary optimization (QUBO) forms. We evaluate the proposed method using two real-world datasets including Sapporo and Tokyo, Japan. Experimental results show that the MTPP can be effectively solved using Ising machines compared with the conventional methods in terms of the solution quality and the execution time. To the best of our knowledge, this study is the first solution of the MTPP using Ising machines.

© 2020 IEEE. Existing studies focus on text information while ignoring category and data source information, both of which are verified to be important in interpreting sentiments in travel comments in this paper. Furthermore, the unique linguistic characteristics of Japanese cause difficulty in applying the conventional token-based word segmentation methods to Japanese comments directly. In this paper, we propose a method of stem-based segmentation based on Japanese linguistic characteristics and incorporate it with category and data source information into a hierarchical network model for document-level sentiment classification. Empirical results of our proposed model outperform existing models on a real-world dataset.

© Springer Nature Switzerland AG 2020. To help people smoothly and efficiently make travel decisions, we utilize the advantages of travel comments posted by thousands of other travelers. In this paper, we analyze the feasibility of exploring landmark activity queries and representative examples from Japanese travel comments. Contributions in this paper include a framework for extracting activity concerned keywords and queries, quantifying the relationship between landmark activities and comment contents. An evaluation of activity-example extraction is conducted in two case studies through 18,939 travel comments.

© 2018 IEEE. Online media communities have globally spanned and have increasingly accelerated the development of intelligent travel recommendation systems in both academic and industrial fields. However, there is a bottleneck that differences in users' seasonal travel distributions (when to visit) in various language groups are ignored. This paper proposes a seasonal activity prediction algorithm based on user comments over the period of 2012 to 2017 in different language groups. We take the advantage of online user comments which provide visiting time for each landmark and detailed activity description. With the accumulation of 417,787 user comments on TripAdvisor for 300 landmarks in three big cities, we analyze the language-specific differences in travel distributions. After that, prediction of future travel distribution for each language group is generated. Then potential peak and off seasons of each landmark are distinguished and representative seasonal activities are extracted through comment contents for peak and off seasons, respectively. Experimental results in the three cities show that the proposed algorithm is more accurate in terms of peak season detection and seasonal activity prediction than previous studies.

© 2019, Springer Nature Switzerland AG. This paper proposes a personalized landmark recommendation algorithm aiming at exploring new sights into the determinants of landmark satisfaction prediction. We gather 1,219,048 user-generated comments in Tokyo, Shanghai and New York from four travel websites. We find that users have diverse satisfaction on landmarks those findings, we propose an effective algorithm for personalize landmark satisfaction prediction. Our algorithm provides the top-6 landmarks with the highest satisfaction to users for a one-day trip plan our proposed algorithm has better performances than previous studies from the viewpoints of landmark recommendation and landmark satisfaction prediction.

© 2018 IEEE. This paper proposes a personalized landmark recommendation algorithm based on the prediction of users' satisfaction on landmarks. We have accumulated 270,239 user-generated comments from travel websites of Ctrip, Jaran and TripAdvisor for 196 landmarks in Tokyo, Japan. We find that users do have different satisfaction on landmarks depending on their commonly used languages and travel websites. Then we establish a database for landmarks with abundant and accurate landmark type and landmark satisfaction information. Finally, we propose an effective personalized landmark satisfaction prediction algorithm based on users' landmark type, language and travel website preferences. After that, landmarks with the top-6 highest satisfaction are provided to the user for a one-day visit plan in Tokyo. Experimental results demonstrate that the proposed algorithm can recommend landmarks that fit the user's preferences and our algorithm also successfully predicts the user's landmark satisfaction with a low error rate less than 7%, which is superior to other previous studies.

© 2018 IEEE. This paper proposes a road-illuminance level inference method based on the naive Bayesian analysis. We investigate quantities and types of road lights and landmarks with a large set of roads in real environments and reorganize them into two safety classes, safe or unsafe, with seven road attributes. Then we carry out data learning using three types of datasets according to different groups of the road attributes. Experimental results demonstrate that the proposed method successfully classifies a set of roads with seven attributes into safe ones and unsafe ones with the accuracy of more than 85%, which is superior to other machine-learning based methods and a manual-based method.

© 2017 IEEE. Travel route recommendation can strongly influence users' satisfaction and the success of touristic businesses. This paper proposes a personalized travel recommendation algorithm with time planning. We use landmark categorization and region clustering to obtain effective elements. Then we build a travel map to generate all possible travel routes. Our proposed algorithm has higher precision in landmark recommendation and time planning than thoes in previous algorithms.

Copyright © 2017 The Institute of Electronics, Information and Communication Engineers. In this paper, we propose a safe and comprehensive route finding algorithm for pedestrians based on lighting and landmark conditions. Safety and comprehensiveness can be predicted by the five possible indicators: (1) lighting conditions, (2) landmark visibility, (3) landmark effectiveness, (4) turning counts along a route, and (5) road widths. We first investigate impacts of these five indicators on pedestrians' perceptions on safety and comprehensiveness during route findings. After that, a route finding algorithm is proposed for pedestrians. In the algorithm, we design the score based on the indicators (1), (2), (3), and (5) above and also introduce a turning count reduction strategy for the indicator (4). Thus we find out a safe and comprehensive route through them. In particular, we design daytime score and nighttime score differently and find out an appropriate route depending on the time periods. Experimental simulation results demonstrate that the proposed algorithm obtains higher scores compared to several existing algorithms. We also demonstrate that the proposed algorithm is able to find out safe and comprehensive routes for pedestrians in real environments in accordance with questionnaire results.

© 2016 IEEE. This paper proposes a safe and comprehensive route finding method for pedestrians. We evaluate five factors that do relieve pedestrians' fear of darkness. Based upon the evaluation, we propose a comprehensive route finding method by taking road width and reduction on turning points into consideration. The experimental results on real outdoor environments under different lighting situations confirm that the proposed method can obtain safety and comprehensive routes for pedestrians.

© 2015 IEEE. This paper proposes a landmark-based route recommendation method for enjoyable walking atmosphere strategies by accumulating and analyzing geographical information. We utilize landmark categorization and region clustering to obtain effective elements. Experimental results demonstrate that our proposed method improves walking environment quality and confirm that it is applicable in both urban and rural areas.