| 徳田 博行 | M, 2回目発表 | ユビキタスコンピューティングシステム | 安本 慶一, 飯田 元, 荒川 豊, 諏訪 博彦 |
|
title: Continuous Micro-Learning Support based on Context-Aware Interruption and Searching Words
abstract: A review is an important process to keep remembering something such as language learning. However, it is hard to make a time to review continuously. Therefore, we propose a system that supports continuous and effective review. It asks a user about past-learned words at an appropriate timing via push notification. We call it a Context-Aware Interruption. Our system prompts to use up a short free moment and review the words which user should remember in our life for a quick review. language of the presentation: Japanese | |||
| 真鍋 優 | M, 2回目発表 | 大規模システム管理 | 笠原 正治, 飯田 元, 笹部 昌弘, 川原 純 |
|
title: On designing an auction resource allocation model in Hadoop datacenter
abstract: Hadoop is an open source software framework for distributed file systems and parallel processing. It allows users to handle a massive amount of data with a large computing cluster. To avoid spending much costs of storing big data and operating such clusters, users desire to share them. However, it is assumed that default schedulers, such as a FIFO scheduler and Fair scheduler, are used by a small number of users at one time. In previous research, a scheduler called Dynamic Proportional Share Scheduling was proposed. It allows many users to simultaneously use a large computing cluster. Users of a cluster enter a bid for using resources and the scheduler allocates the resources to them in proportional to the amounts of their bids. We propose a new method to determine the price for using resources so that users actually spend the cost it is acceptable for them to pay. In our method, we used an auction mechanism with reservation prices for resource allocation and price setting. By using the reservation prices, it is possible to prohibit bids below a certain amount and guarantee the administrator's revenues. We compare this method to the one in previous research in terms of execution time and administrator's revenues. The result shows that the revenues increased by 8% to 34% from previous research. language of the presentation: Japanese | |||
| 坂本 大生 | M, 2回目発表 | 環境知能学 | 萩田 紀博, 小笠原 司, 神原 誠之 |
|
title: ”Kawaii” Pose Modeling in Robot
abstract: Intimacy is considered to be important for continuous use of robots. Also, in the field of psychology, there are data that people feel more intimate when they see cute things. Therefore, there is research to improve the intimacy by cute robot's voice and form. However, despite being able to pose freely, the robot has few studies on cute poses. So, letting the robot take a cute pose to make it more intimate, I aim to realize an environment that can continuously use the robot. In this research, we derive common properties for poses that people feel cute. language of the presentation: Japanese | |||
| 松本 愼太郎 | M, 2回目発表 | 環境知能学 | 萩田 紀博, 小笠原 司, 神原 誠之 |
|
title: Collecting and Analyzing Data of Vehicle Behaviors, Environment, and Passengers for Estimating Autonomous Vehicle Stress
abstract: Passengers in the autonomous driving space have anxiety and fear to the autonomous driving system, due to factors such as vehicle behaviors different from his ideal and obstacles in the surrounding environment of the vehicle. In order to eliminate such factors and enable continued use of the autonomous driving vehicle, close interactions called "Comfort Intelligence(CI)" is required between the autonomous driving vehicles and the its passengers. In this research, the purpose is to estimate the factor of anxiety and fear felt by passengers of an autonomous driving vehicle called "Autonomous Vehicle Stress(AVS)". For estimating AVS factors, we collect and analyze data of vehicle behavior, vehicle external environment, and passengers by subject experiments using real cars. language of the presentation: Japanese | |||
| 坂田 大直 | M, 2回目発表 | 環境知能学 | 萩田 紀博, 加藤 博一, 神原 誠之 |
|
title: How Robots Should Behave to Make Interlocutors' Utterences Increased in Telecommunication via Robots.
abstract: The number of single-living elderly people is increasing. It is often difficult for them to meet new people. Previous research demonstrated that people feel easy to do self-disclosure via telecommunication tools. However, it is still difficult for not-talkative to communivate with others. I'm working on how robots should behave in this situation. language of the presentation: Japanese | |||
| 津﨑 圭祐 | M, 2回目発表 | 環境知能学 | 萩田 紀博, 加藤 博一, 神原 誠之 |
|
title:
Effect Investigation of dialogue motivation by contagious laughter of interaction robot during TV watching
abstract:This research investigates effect of dialogue motivation by contagious laughter of interactive robot during TV watching. Conventional research in psychonomic field has shown that contagious laughter has sufficient stimulus for laughs and smile for others. This study carries out experiments which evaluate psychologic influence of user when the interactive robots give user contagious laughs. This presentation shows experimental environments which the interactive robots give contagious laughs to user during TV watching.
language of the presentation: Japanese | |||
| 土田 将司 | D, 中間発表 | ディペンダブルシステム学 | 井上 美智子, 笠原 正治, 大下 福仁, 新谷 道広 |
|
title: Mobile agent algorithms with Byzantine fault tolerance
abstract: In recent years, a distributed system consisting of multiple computers (nodes) has become larger. Each node communicates with each other, and communication complexity increases in huge distributed system. This makes it complicated to design distributed systems because developers must maintain a huge number of nodes and treat massive date communication among them. To alleviate these problems, mobile agents have attracted attention as a new paradigm. Agents are software programs that can autonomously move from a node to a node and execute various tasks in distributed systems. By using agents, nodes do not need to communicate with each other because agents themselves can collect and analyze data by moving around distributed systems. Since a node exists as a physical entity, it may crash. In addition, some of agents might be crashed because they move on the distributed systems and might be affected by several noise. Crashed agents do not follow the algorithm and behaves arbitrarily. We call such agents Byzantine agents. In this work, I try to propose algorithms which is not influenced by the behavior of Byzantine agents. language of the presentation: Japanese | |||
| 藤谷 拓矢 | M, 2回目発表 | ディペンダブルシステム学 | 井上 美智子, 笠原 正治, 大下 福仁, 新谷 道広 |
|
title: Evacuation of robots from isosceles triangle in face to face communication model
abstract: It is rare case that you knew the route of evacuation when evacuating in case of a fire or other emergency. So,The time required for evacuation should include the time to look for an exit. When thinking about evacuation from a certain area, previous study considered the area as a circle which include the area and considered the time necessary for evacuation. In this study, The area is true area. (we estimate the time necessary for evacuation on condition that the area is isosceles triangle) we consider the evacuation of robots from isosceles triangle in face to face communication model and show lower bound, evacuation algorithm and upper bound. language of the presentation : Japanese | |||
| ZHU ZEYU | M, 2回目発表 | ネットワークシステム学 | 岡田 実, 笠原 正治, 東野 武史, Duong Quang Thang |
| 上田 一貴 | M, 2回目発表 | ネットワークシステム学 | 岡田 実, 笠原 正治, 東野 武史, Duong Quang Thang |
|
title: Integration of communication and power supply using leaky coaxial cable antenna (LCX)
abstract: Recently, a leaky coaxial cable (hereinafter referred to as LCX: Leaky coaxial cable) antenna that can radiate radio waves of 5 GHz band and has a diameter of about 5 mm is emerging.A radio communication service such as wireless LAN is called LCX antenna It has been shown that high frequency utilization efficiency is achieved within the area expanded by. Therefore, I show that the existing leaky coaxial cable antenna can be used as a power transmitting antenna for wireless power supply. language of the presentation: Japanese | |||