It has been established that the second-order stochastic gradient descent (2SGD) method can potentially achieve generalization performance as well as empirical optimum in a single pass (i.e., epoch) through the training examples. However, 2SGD requires computing the inverse of the Hessian matrix of the loss function, which is prohibitively expensive. This paper presents Periodic Step-size Adaptation (PSA), which approximates the Jacobian matrix of the mapping function and explores a linear relation between the Jacobian and Hessian to approximate the Hessian periodically and achieve near-optimal results in experiments on a wide variety of models and tasks.

Introduction In the last years, many useful NLP tools have been developed and many of them are now even available commercially. Most of these tools are monolingual or multi-monolingual, meaning that the software can deal with more than one language, but that the re sults will al ways be displayed in the same language as the text. We therefore distinguish these applications from cross-lingual software, which is software that helps to transgress the language bound ary. Examples for such applications are machine translation and cross-lingual document retrieval, i.e. retrieval using search engines which allow to en ter a search term in one language and which also yield results in other languages, usually because the query is translated in one way or another. In our eyes, cross-lingual applications are currently the bottleneck of available NLP tools. To our knowledge, there are no applications that allow comparing documents written in dif ferent languages with each other and there are very few which give users a quick overview of the ap proximate contents of documents written in different languages.

This paper presents the design of an associative memory with feedback that is capable of on-line temporal sequence learning. A framework for on-line sequence learning has been proposed, and different sequence learning models have been analysed according to this framework. The network model is an associative memory with a separate store for the sequence context of a symbol. A sparse distributed memory is used to gain scalability. The context store combines the functionality of a neural layer with a shift register. The sensitivity of the machine to the sequence context is controllable, resulting in different characteristic behaviours. The model can store and predict on-line sequences of various types and length. Numerical simulations on the model have been carried out to determine its properties.

This paper belongs to the area of learning theory that has been variously referred to as prediction with expert advice, competitive on-line prediction, p rediction of individual sequences, and universal on-line learning; see [7] for a re view. There are many proof techniques known in this field; this paper is based on K alnishkan and Vyugin's Weak Aggregating Algorithm [16], but it is possible that som e of the numerous other techniques could be used instead. In Section 2 we give the main definitions and state our main results, Th e-orems 1-4; their proofs are given in Sections 3-6. In Section 7 we inf ormally discuss the notion of stationarity, and Section 8 concludes.

This paper describes some functionalities and features of the Maryland Virtual Patient (MVP) environment. MVP models the process of disease progression, diagnosis and treatment in virtual patients who are endowed with a “body,” a simulation of their physiological and pathological processes, and a “mind,” a set of capabilities of perception, reasoning and action that allow the virtual patient to exhibit independent behavior, participate in a natural language dialog, remember events, hold beliefs about other agents and about specific object and event instances, make decisions and learn.

This paper describes the DynaLearn project, which seeks to address contemporary problems in science education by integrating well established, but currently independent technological developments, and utilize the added value that emerges. Specifically, diagrammatic representations are used for learners to articulate, analyse and communicate ideas, and thereby construct their conceptual knowledge. Ontology mapping is used to find and match co-learners working on similar ideas to provide individualised and mutually benefiting learning opportunities. Virtual characters are used to make the interaction engaging and motivating. The development of the workbench is tuned to fit key topics from environmental science curricula, and evaluated and further improved in the context of existing curricula using case studies. Through this approach, the DynaLearn project will deliver an individualised and engaging cognitive tool for acquiring conceptual knowledge that fits the true nature of this expertise.

We have developed computer environments that support learning by teaching and the use of self regulated learning (SRL) skills through interactions with virtual agents. More specifically, students teach a computer agent, Betty, and can monitor her progress by asking her questions and getting her to take quizzes. The system provides SRL support via dialog-embedded prompts by Betty, the teachable agent, and Mr. Davis, the mentor agent. Our primary goals have been to support learning in complex science domains and facilitate development of metacognitive skills. More recently, we have also employed sequence analysis schemes and hidden Markov model (HMM) methods for assigning context to and deriving aggregated student behavior sequences from activity data. These techniques allow us to go beyond analyses of individual behaviors, instead examining how these behaviors cohere in larger patterns. We discuss the information derived from these models, and draw inferences on students’ use of self-regulated learning strategies.

The goal of this paper is to present four key assumptions regarding the measurement of cognitive and metacognitive regulatory processes used during learning with hypermedia. First, we assume it is possible to detect, trace, model, and foster SRL processes during learning with hypermedia. Second, understanding the complex nature of the regulatory processes during learning with hypermedia is critical in determining why certain processes are used throughout a learning task. Third, it is assumed that the use of SRL processes can dynamically change over time and that they are cyclical in nature (influenced by internal and external conditions and feedback mechanisms). Fourth, capturing, identifying, and classifying SRL processes used during learning with hypermedia is a rather challenging task.

Despite the fact that AI has always been adventurous in trying to elucidate complex aspects of human behaviour, only recently has there been research into computational modelling of humor. One obstacle to progress is the lack of a precise and detailed theory of how humor operates. Nevertheless, since the early 1990s, there have been a number of small programs that create simple verbal humor, and more recently there have been studies of the automatic classification of the humorous status of texts. In addition, there are a number of advocates of the practical uses of computational humor: in user-interfaces, in education, and in advertising. Computer-generated humor is still quite basic, but it could be viewed as a form of exploratory creativity. For computational humor to improve, some hard problems in AI will have to be addressed.

The difference between comedians and their audience is a matter not of kind, but of degree, a difference that is reflected in the vocational emphasis they place on humor. Researchers in the field of computational creativity find themselves in a similar situation. As a subdiscipline of artificial intelligence, computational creativity explores theories and practices that give rise to a phenomenon, creativity, that all intelligent systems, human or machine, can legitimately lay claim to. Who is to say that a given AI system is not creative, insofar as it solves nontrivial problems or generates useful outputs that are not hard wired into its programming? As with comedians' being funny, the difference between studying computational creativity and studying artificial intelligence is one of emphasis rather than one of kind: the field of computational creativity, as typified by a long-running series of workshops at AIrelated conferences, places a vocational emphasis on creativity and attempts to draw together the commonalities of what