In recent years, various studies have investigated intelligent robots that can smoothly communicate with humans. Communication between humans is often achieved through conversation, so it is likely that robots will ne...
详细信息
In recent years, various studies have investigated intelligent robots that can smoothly communicate with humans. Communication between humans is often achieved through conversation, so it is likely that robots will need to be able to perform human-like conversation. There are various types of human conversation. Intelligent robots will be expected to perform conversations a lot as human. In order to realize such conversation in robots, resourses for making new topics is required. We herein focus on the news website headlines because headlines are a more appropriate format for initiating conversation than the text of articles. However, there is a problem in that specific information is lacking from headlines. Therefore, this paper present a method of embodying the meaning of headlines using news articles for using headlines as a resource of the robot conversation. The proposed method embodies the meanings of news headlines by adding or replacing words in the headline with words from the articles. As a result, the proposed method was able to embody the meanings of Japanese headlines in a natural manner for approximately 58.3% of the 120 headlines belonging to the test set.
Humans communicate with others using natural language because many expressions in natural language can convey the same message, humans interpret these expressions flexibly based on their knowledge of words and associa...
详细信息
Humans communicate with others using natural language because many expressions in natural language can convey the same message, humans interpret these expressions flexibly based on their knowledge of words and association skills. An Association System was constructed on a computer by applying Concept Bases and the degree of association. This paper proposes a method of generating association word from multiple words based on Association System to show that it can achieve humanlike associative abilities on a computer. The proposed method generated a natural human association with 61.0% accuracy and 77.0% recall.
Introduction: With the range of educational tools available it is now realistic for learner models to take account of broader information, and there are strong arguments for placing open learner models in the centre o...
详细信息
Automatic recognition of emotional states from human speech is a current research topic with a wide range. In this paper an attempt has been made to recognize and classify the speech emotion from three language databa...
详细信息
Automatic recognition of emotional states from human speech is a current research topic with a wide range. In this paper an attempt has been made to recognize and classify the speech emotion from three language databases, namely, Berlin, Japan and Thai emotion databases. Speech features consisting of Fundamental Frequency (F0), Energy, Zero Crossing Rate (ZCR), Linear Predictive Coding (LPC) and Mel Frequency Cepstral Coefficient (MFCC) from short-time wavelet signals are comprehensively investigated. In this regard, Support Vector Machines (SVM) is utilized as the classification model. Empirical experimentation shows that the combined features of F0, Energy and MFCC provide the highest accuracy on all databases provided using the linear kernel. It gives 89.80%, 93.57% and 98.00% classification accuracy for Berlin, Japan and Thai emotions databases, respectively.
We study Monte Carlo tree search (MCTS) in zero-sum extensive-form games with perfect information and simultaneous moves. We present a general template of MCTS algorithms for these games, which can be instantiated by ...
详细信息
ISBN:
(纸本)9781632660244
We study Monte Carlo tree search (MCTS) in zero-sum extensive-form games with perfect information and simultaneous moves. We present a general template of MCTS algorithms for these games, which can be instantiated by various selection methods. We formally prove that if a selection method is e-Hannan consistent in a matrix game and satisfies additional requirements on exploration, then the MCTS algorithm eventually converges to an approximate Nash equilibrium (NE) of the extensive-form game. We empirically evaluate this claim using regret matching and Exp3 as the selection methods on randomly generated games and empirically selected worst case games. We confirm the formal result and show that additional MCTS variants also converge to approximate NE on the evaluated games.
Social exergames provide immersive experiences of social interaction via online multiplayer games, ranging from simple group exercises (e.g., virtual cycling/rowing) to more structured multiplayer games (e.g., coopera...
详细信息
With increasing adoption and presence of Web services, designing novel approaches for efficient Web services recommendation has become steadily more important. Existing Web services discovery and recommendation approa...
详细信息
With increasing adoption and presence of Web services, designing novel approaches for efficient Web services recommendation has become steadily more important. Existing Web services discovery and recommendation approaches focus on either perishing UDDI registries, or keyword-dominant Web service search engines, which possess many limitations such as insufficient recommendation performance and heavy dependence on the input from users such as preparing complicated queries. In this paper, we propose a novel approach that dynamically recommends Web services that fit users' interests. Our approach is a hybrid one in the sense that it combines collaborative filtering and content-based recommendation. In particular, our approach considers simultaneously both rating data and content data of Web services using a three-way aspect model. Unobservable user preferences are represented by introducing a set of latent variables, which is statistically estimated. To verify the proposed approach, we conduct experiments using 3, 693 real-world Web services. The experimental results show that our approach outperforms the two conventional methods on recommendation performance.
This panel explores the many roles of data analytics in today's cross-domain collaborations. In some instances, cross-domain analytics are required to understand big data. In others, big data holds the key to unde...
This panel explores the many roles of data analytics in today's cross-domain collaborations. In some instances, cross-domain analytics are required to understand big data. In others, big data holds the key to understanding and evaluating how people collaborate across domains. Panelists will present their experiences with big data and collaboration, and discussion will be guided by the following questions: How does “understanding data through collaboration” relate to “understanding collaboration through data”? What modeling tools, visualizations, abstractions, and metrics can help facilitate collaborative discovery from big data? To what extent should these tools be tailored to specific domains or groups? How much do we need to know about another domain or group before we can intersect domain specific activities that constitute the context to various analytical processes, and cross-pollinate ideas for deeper analytics? What optimally would that other domain or group need to know about us? How much of this knowledge can we extract from data?
The new trend in the process of data-intensive management indicates the importance of a distributed file system for both Internet large scale services and cloud computing environments. I/O latency and application buff...
详细信息
ISBN:
(纸本)9781479904051
The new trend in the process of data-intensive management indicates the importance of a distributed file system for both Internet large scale services and cloud computing environments. I/O latency and application buffering sizes are two of a number of issues that are essential to be analysed on different class of distributed file systems. In this paper, it is presented a research work comparing four different high performance distributed file systems. Those systems were employed to support a medical image server application in a private storage environment. Experimental results highlight the importance of an appropriate distributed file system to provide a differential level of performance considering application specific characteristics.
暂无评论