Abstract: Advances in the publishing world have emerged new models of digital library development. Open access publishing modes are expanding their presence and realize the digital library idea in various means. While user-centered evaluation of digital libraries has drawn considerable attention during the last years, these systems are currently viewed from the publishing, economic and scientometric perspectives. The present study explores the concepts of usefulness and usability in the evaluation of an e-print archive. The results demonstrate that several attributes of usefulness, such as the level and the relevance of information, and usability, such as easiness of use and learnability, as well as functionalities commonly met in these systems, affect user interaction and satisfaction.

Abstract: In this paper, I show that a constraint-based approach to DOM, such as the OT system proposed by Aissen 2003, leads to the conclusion that the child will entertain a number of grammars different from the target grammar, before acquiring the final ranking of constraints, even under the learnability assumptions of the OT framework (such as those made by Tesar and Smolensky 1998, 2000). The data examined here, from Spanish-speaking children in the CHILDES database, clearly shows that children master Spanish DOM with a performance virtually errorless. This raises doubts regarding the capacity of the OT framework to explain a key aspect of human language, namely, the process of acquisition.

Abstract: In this paper, I show that a constraint-based approach to DOM, such as the OT system proposed by Aissen 2003, leads to the conclusion that the child will entertain a number of grammars different from the target grammar, before acquiring the final ranking of constraints, even under the learnability assumptions of the OT framework (such as those made by Tesar and Smolensky 1998, 2000). The data examined here, from Spanish-speaking children in the CHILDES database, clearly shows that children master Spanish DOM with a performance virtually errorless. This raises doubts regarding the capacity of the OT framework to explain a key aspect of human language, namely, the process of acquisition.

Abstract: The present text is a re-edition of Volume I of Formal Grammars in Linguistics and Psycholinguistics , a three-volume work published in 1974. This volume is an entirely self-contained introduction to the theory of formal grammars and automata, which hasn’t lost any of its relevance. Of course, major new developments have seen the light since this introduction was first published, but it still provides the indispensible basic notions from which later work proceeded. The author’s reasons for writing this text are still relevant: an introduction that does not suppose an acquaintance with sophisticated mathematical theories and methods, that is intended specifically for linguists and psycholinguists (thus including such topics as learnability and probabilistic grammars), and that provides students of language with a reference text for the basic notions in the theory of formal grammars and automata, as they keep being referred to in linguistic and psycholinguistic publications; the subject index of this introduction can be used to find definitions of a wide range of technical terms. An appendix has been added with further references to some of the core new developments since this book originally appeared.

Abstract: Since the introduction of rational expectations, there have been issues with multiple equilibria and equilibrium selection. We study the connections between determinacy of rational expectations equilibrium and learnability of that equilibrium in a general class of purely forward-looking models. Our framework is sufficiently flexible to encompass lags in agents' information and either finite horizon or infinite horizon approaches to learning. We are able to isolate conditions under which determinacy does and does not imply learnability and also conditions under which long-horizon forecasts make a clear difference for learnability. Finally, we apply our result to a relatively general New Keynesian model. [ABSTRACT FROM AUTHOR

Abstract: The aim of this paper is to examine the learning processes that subjects undertake when they start using gaze as computer input. A 7-day experiment with eight Japanese students was carried out to record novice users' eye movement data during typing of 110 sentences. The experiment revealed that inefficient eye movements was dramatically reduced after only 15-25 sentences of typing, equal to approximately 3-4 h of practice. The performance data fits a general learning model based on the power law of practice. The learning model can be used to estimate further improvements in gaze typing performance. Our experimental results encourage the use of gaze-based interfaces for severely disabled people. This paper provides a taxonomy for gaze actions on dwell time-activated keys and it presents a method by which the learnability of gaze interfaces may be documented.

Abstract: The formal study of intelligence has largely focused on learning and reasoning, the processes by which knowledge is, respectively, acquired and applied. This dissertation investigates how the two processes may be undertaken together in an autodidactic, or self-taught, manner. The thesis put forward is that the development of such a unified framework rests on the principled understanding of a third process, that of sensing. Sensing is formalized in this dissertation as the process by which some underlying reality completely specifying a state of affairs is mapped to an appearance explicitly offering only partial information. Learning is employed to discover the structure of the reality, and reasoning is employed to recover as much of the missing information as possible. Emphasis is placed on the tractability of learning and reasoning, and on the existence of formal guarantees on the accuracy of the information recovered, making only minimal assumptions on the nature of information loss during the sensing phase. An investigation of the conditions under which the task of information recovery is feasible is undertaken. It is shown that it suffices, and is optimal in some precisely defined sense, to induce rules that are simply consistent with the observed appearances. For environments with structure expressible via monotone rules, learning consistently from partial appearances reduces to learning from complete appearances, allowing for known positive results to be lifted to the case of autodidactic learning. On the negative side, there exist environments where partial appearances compromise learnability. The contribution of chaining rules—induced or externally provided ones—for information recovery is then examined, and is shown to be that of increasing the combined predictive soundness and completeness. This result provides apparently the first formal separation between multi-layered and single-layered reasoning in this context. It is further established that the learning and reasoning processes cannot be completely decoupled in the autodidactic setting. Instead, an approach that interleaves the two processes is introduced, which proceeds by learning the rules to be employed for multi-layered reasoning in an iterative manner, one layer at a time. This approach of employing interim reasoning, or reasoning while learning, is shown to suffice and to be a universal approach for the induction of knowledge that is to be reasoned with. The design and implementation of a system for automatically acquiring and manipulating knowledge is finally considered. Semantic information extracted from a natural language text corpus is interpreted, following the theory, as partial information about the real world. It is argued that rules induced from such information capture some commonsense knowledge. This knowledge is subsequently employed to recover information that is not explicitly stated in the corpus. Experiments were performed on a massive scale, and serious computational challenges had to be addressed to ensure scalability. The experimental setting was designed with the novel goal of detecting whether commonsense knowledge has been extracted. The experimental results presented suggest that this goal has been achieved to a measurable degree.

Abstract: devices have much potential to support older adults in their daily lives. However older adults have reported that they find mobile devices, such as existing mobile phones, difficult to learn to use. We argue that more research is needed to improve the learnability of mobile software applications for older adults. We present three different approaches for improving mobile device learnability: improving graphical icons, using multi- layered interfaces, and augmenting the mobile interface.

Abstract: It is investigated for which choice of a parameter q, denoting the number of contexts, the class of simple external contextual languages is iteratively learnable. On one hand, the class admits, for all values of q, polynomial time learnability provided an adequate choice of the hypothesis space is given. On the other hand, additional constraints like consistency and conservativeness or the use of a one-one hypothesis space changes the picture -- iterative learning limits the long term memory of the learner to the current hypothesis and these constraints further hinder storage of information via padding of this hypothesis. It is shown that if q> 3, then simple external contextual languages are not iteratively learnable using a class preserving one-one hypothesis space, while for q= 1 it is iteratively learnable, even in polynomial time. For the intermediate levels, there is some indication that iterative learnability using a class preserving one-one hypothesis space might depend on the size of the alphabet. It is also investigated for which choice of the parameters, the simple external contextual languages can be learnt by a consistent and conservative iterative learner.

Abstract: Empirical grammatical inference systems are practical systems that learn structure from sequences, in contrast to theoretical grammatical inference systems, which prove learnability of certain classes of grammars. All current empirical grammatical inference evaluation methods are problematic, i.e. dependency on language experts, appropriateness and quality of an underlying grammar of the data, and influence of the parameters of the evaluation metrics. Here, we propose a modification of an evaluation method to reduce the ambiguity of results.

Abstract: It has been widely recognized that learning and evolution have the commonality of involving adaptive processes that once started do not need a programmer or designer. It is tempting to seek some mystical extra power in evolution, beyond that of learning, simply because of the apparently spectacular consequences of evolution that we see around us. However, such approaches have not succeeded to date. In response to this situation one of the authors made the, apparently radical, suggestion that evolution is nothing other than a constrained form of computational learning. In [Val08] a notion of evolvability was defined in a similar spirit to the definition of learnability. The goal of the definition is to offer a rigorous basis for the analysis of evolution and for distinguishing between efficient evolution and evolution that is only realized in some exponentially far limit. Before summarizing this framework we describe the following motivating concrete instance. Consider the 20,000 or so genes in the human genome. For each such gene the condition under which the protein corresponding to it is expressed, in terms of all the other proteins, is encoded in its regulatory region. In other words each of the 20,000 or so proteins is controlled by a function f of the other 20,000 or so proteins. The issue here is that if the function f is restricted to too small a class then it will not be expressive enough to perform the complex functions of biology. On the other hand, if the function is an arbitrary function, or from a too extensive a class, then no evolutionary algorithm will exist to maintain the viability of this genetic network of functions as environmental conditions change. The goal of this evolvability theory is, among other things, to understand how broad and expressive these functions can be allowed to be while still permitting their efficient evolution. The following is an abbreviated summary of the basic definitional framework. Let X = {0, 1}n be an n-dimensional space of experiences or examples (e.g. in the above instance the expression levels of the proteins), a set C of functions (e.g. the functions by which the expression level of each protein is determined in terms of the expression levels of the others), and a set R of representations of functions (e.g. the DNA strings of the genes). Also we define an ideal function f , which would define for each vector x ∈ X the best value from the viewpoint of the evolving organism. In the current instance, for each combination of expression levels

Abstract: The purpose of the study is to examine the usability features and individual acceptance of e-government applications of the three Turkish social security institutions. An experimental study with 45 adults revealed information on the interaction between usability features (learnability, memorability, efficiency, errors and satisfaction) and user acceptance (perceived usefulness and perceived ease of use) of those applications. The tests revealed significant relationships between constructs of usability and acceptance of those e-government applications. Further analysis illustrated that the participants' education level, previous usage of the sites, experience in computer and internet use and frequency of internet usage affected the accomplishment of the given tasks. In the applications, the participants commonly experienced problems in finding a relevant link to the task in hand and in using the interfaces. The study suggests that the e-government applications should include reusable learning objects to instruct use of the applications in order to promote those applications.

Abstract: This paper deals with quality enhancement of e-learning activities through the satisfaction analysis expressed by university learners while using specific learning modules. The paper considers a newly engineered survey that evaluates not only quality in use but also quality in learning and learner involvement. The quality of the interface, the friendliness of the contents, the accessibility of concepts and the capability to provide new knowledge and skills are all considered. Through the paper, the various aspects concerning the different discriminating factors that characterize e-learning processes are investigated. The results, obtained during university teaching activities, show the usefulness of the survey in individualizing the more improvable factors by understanding effective learners’ needing. Key-Words: e-learning, learnability, quality evaluation, survey, usability.

Abstract: We show that artificial language evolution involves the interplay of two opposing forces: pressure toward simple representations imposed by the dynamics of collective learning, and pressure towards complex representations imposed by requirements of agents’ tasks. The push-pull of these two forces results in the emergence of a language that is balanced: “simple but not too simple.” We introduce the classification game to study the emergence of these balanced languages and their properties. Our agents use artificial neural networks to learn how to solve tasks, and a simple counting algorithm to simultaneously learn a language as a form-meaning mapping. We show that task-language coupling drives the simplicity-complexity balance, and that both compositional and holistic languages can emerge.

Abstract: A brief timetable of language development is outlined. This article then deals in turn with the major aspects of language learning in terms of infancy, learning words, learning morphology, early grammar, later grammar, the learning of pragmatic and metalinguistic skills and, finally, some brief reflections on atypical development. The relevant theoretical issues are covered as they arise in each section. They are taken up again in the last section on learnability and constituency.

Abstract: In a very broad class of dynamic linear models, if agents possess knowledge of current endogenous variables in a least-squares learning process, determinacy of a rational expectations (RE) equilibrium is sufficient but not necessary for learnability of that equilibrium. Thus, since learnability is an attractive necessary condition for plausibility of any equilibrium, there may exist a single plausible RE solution even in cases of indeterminacy. This paper proposes and outlines a distinct criterion that plausible models should possess, termed "well formulated" (WF), which rules out infinite discontinuities in the implied impulse response functions. The paper explores the relationship between this WF property and learnability, under the information assumption mentioned above, and finds that they often agree but neither strictly implies the other. Extending the P-matrix requirement, implied for specified matrices by the WF property, to one that demands positive dominant-diagonal matrices would guarantee both WF and learnability, but a suitable rationale has not been found. Finally, under a second information assumption, which gives the agents only lagged information on endogenous variables during the learning process, the situation is less favorable in the sense that learnability can be guaranteed only under special assumptions.

Abstract: This paper examines the logical problem of language acquisition drawing upon an experimental study on children’s knowledge of anaphoric one by Lidz, Waxman and Freedman (2003). The finding was that, upon being presented with the instruction “ Look! A yellow bottle. Do you see another one? ”, 18-month-old children prefer to look at a yellow bottle rather than to a bottle of a different color. According to Lidz et al. (2003), the results that children cannot interpret one as anaphoric to head nouns. We point out that the experimental findings are not explained under the hypothesis offered by the authors of that study. Secondly, we consider whether, under current assumptions, children’s knowledge of anaphoric one can be inferred from the properties of the final state. Thirdly, we reaffirm the validity of the Poverty of the Stimulus argument, despite the challenge posed by the learning model proposed by Regier and Gahl (2004). Finally, we draw upon recent psycholinguistic work to propose an explanation for the findings documented by Lidz et al. (2003) that is independent from – though consistent with – their knowledge of the constraint on anaphoric one .

Abstract: The purpose of this symposium is to move beyond speculations about how knowledge about complex systems might be important for students to understand to focus on empirical research into the learnability of these ideas. For example, do complex systems ideas represent learning challenges that are qualitatively different than learning other scientific knowledge? What are the differences in pre-conceptions students have about complex systems phenomena and more expert scientific ways of thinking in these areas? What are the profiles of successful and less successful ways of learning about complex systems conceptual perspectives? Can complex systems provide conceptual perspectives for cognitively "seeing" physical and social sciences subjects in new and interconnected ways? It is hoped the papers in this session will provide insights into these questions and other theoretical and research issues in the learning sciences.

Abstract: Selection through iterated learning explains no more than other non-functional accounts, such as universal grammar, why language is so well-designed for communicative efficiency. It does not predict several distinctive features of language like central embedding, large lexicons or the lack of iconicity, that seem to serve communication purposes at the expense of learnability.

Abstract: The use of Computer Aided Software Engineering (CASE) tools for teaching object-oriented systems analysis and design (OOSAD) and the Unified Modelling Language (UML) has many potential benefits, but there are several problems associated with the usability and learnability of these tools. This paper describes a study undertaken to determine if computing students from a linguistically and technologically diverse community experience problems with learning to use a CASE tool, and to determine if there is a relationship between two user characteristics of the students and the learnability of CASE tools.

Abstract: The importance of understanding the wants and needs of an information source's users is a key belief of user-centered design. In digital libraries, we find two approaches to including users as factors in system design. User studies tend to provide a more abstract view and can be associated with information science research. Usability assessment tends to be associated with the practical assessment of the ease, efficiency, learnability, and flexibility of the digital library. However, usability assessment seldom deals with differences in user communities. This study of the information needs and information seeking values of three groups of digital library users illustrates how important it is to understand differences between user communities, to have a “nuanced view of the user”, to support informed decisions assessing the need for digital library content and organization.

Abstract: Interaction and usability aspects of software systems are critical software quality measures. This paper reports on work done in the European funded wearable computing project wearIT@work. Utilizing user centered design, we present the results of a user study conducted to evaluate different wearable input devices for menu-selection tasks in aircraft maintenance. Due to access to end users provided by wearIT@work, study participants could be recruited from both real maintenance workers and university students to study performance and group differences. Results already exhibit strong group differences. An unfamiliar data glove device using gestures was found to outperform natural speech command interaction. A familiar mouse device exhibited best performance and quick learnability but was judged inappropriate by maintainers for their daily work. Besides suffering from handling problems, speech interaction was found inapplicable for maintainers due to their language ability. Our results suggest and emphasize the necessity to include real end-users in projects to reasonably study industrial wearable computing applications and their constraints.

Abstract: In this paper we investigate the status of three different types of agreement spell-outs in English/Spanish bilingual data: (1) free [+pronominal] agreement morphemes (the nominative pronouns in English: he, they...); (2) bound [+pronominal] agreement morphemes (the Spanish person agreement markers in V: -mos, -s...); and (3) bound [-pronominal] agreement morphemes (the English third person singular marker –s). Using production data from two English/Spanish bilingual children, we will establish interlinguistic contrasts (e.g., English -s spell-out versus Spanish person agreement markers), as well as intralinguistic contrasts (e.g., English -s spell-out versus English pronominal subjects). This will allow us to characterize these different types of agreement spell-outs with respect to learnability. Our specific aim is to determine whether the L1 development of these agreement spell-outs in our production data provides evidence for the different status attributed to them by linguistic theory. The use of Spanish/English bilingual data allows us to analyze the three different types of agreement spell-outs and test our hypotheses within the same ‘population’. We will start by reviewing Alexiadou and Anagnostopoulou (1998) proposal from linguistic theory regarding the distinction between [+null argument] languages and [-null argument] languages. In relation to this analysis, we will discuss the notion of markedness and how the three spell-outs can be accounted for in terms of different markedness proposals. We will then present and analyze our data in terms of productivity and error index. The discussion of the data will be followed by our conclusions and suggestions for further research.

Abstract: In this paper, we report on our investigation into people's ability to use the volume of their voice to create digital drawings. This is especially hopeful for artists with upper limb disabilities who show remarkable endurance, patience and determination to create art with whatever means available to them. We developed a prototype 'voice art' system to investigate the potential of this form of interaction. Our first experiment showed that varying the volume of the voice is both comfortable and intuitive. However, it was ascertained the instant usability of volume based control is better suited for target acquisition tasks as opposed to drawing tasks. Nevertheless, the results indicate the potential learnability of volume control for drawing. A second experiment investigated the long-term learnability of this form of interaction. The results show that drawing using voice volume control is a skill that can be developed with time. We believe that voice volume control has implications beyond assisting artists with upper limb disabilities. Such possible implications may be: an alternative mode of interaction for disabled people to perform tasks other than creating visual art; for people whose hands are busy elsewhere; and as a voice training system for people with speech impairments.