Rankboost has been shown to be an effective algorithm for combining ranks. However, its ability to generalize well and not overfit is directly related to the choice of weak learner, in the sense that regularization of the rank function is due to the regularization properties of its weak learners. We present a regularization property called consistency in preference and confidence that mathemati...

We give a general overview of C U N T , a userfriendly interactive concept-learner, which can be used as a module for Learning Apprentice Systems. CLINT combines several interesting features : it uses domain-knowledge, generates examples, copes with indirect relevance, shifts its bias, recovers from errors and identifies concepts in the limit.

With the increasing importance of technology in our lives, incidences of adults experiencing anxiety about technology (`technophobia') are on the increase. The eld of technology that has advanced most rapidly in the last 40 years is that of computers and microelectronics. Thus, `com-puterphobia' (or`computer anxiety') is a particularly striking example of the eeects of the rapid growth of our t...

When we talk about “potential learners,” it is unhelpful to dissociate them from their environment, more so because the economic well-being of the respective countries in which they live vary. Also, existing national educational provision and economic and social services are essential contributory factors worthy of consideration. More relevant, perhaps, is that the dictates of the national job ...

Recent interpretations of the Adaboost algorithm view it as performing a gradient descent on a potential function. Simply changing the potential function allows one to create new algorithms related to AdaBoost. However, these new algorithms are generally not known to have the formal boosting property. This paper examines the question of which potential functions lead to new algorithms that are ...

The different papers in this special issue all addressed learning with complex tasks. All of these papers reached only partially the expected results. This discussion on possible factors that may explain these unexpected results. A first issue that is questioned is the functionality of the tools in the studies. Secondly, the learners lack of compliance is addressed. It may have been that the le...

A number of interaction researchers have claimed that recasts might be ambiguous to learners; that is, instead of perceiving recasts as containing corrective feedback, learners might see them simply as literal or semantic repetitions without any corrective element (Long, in press; Lyster & Ranta, 1997). This study investigates learners’ interpretations of recasts in interaction. Videotapes of t...

People often recognize same-race faces better than other-race faces. This cross-race effect (CRE) has been proposed to arise in part because learners devote fewer cognitive resources to encode faces of social out-groups. In three experiments, we evaluated whether learners' other-race mnemonic deficits are due to "cognitive disregard" during study and whether this disregard is under metacognitiv...

This study investigates the extent to which L2 learners with different L1 backgrounds are sensitive to phonetic information contained in the visual cues to a novel phonetic contrast (labial/labiodental contrast), and the degree to which this sensitivity can be increased via intensive training. 36 Spanish-L1 and 47 Japanese-L1 learners of English were initially tested on their perception of the ...

This paper investigated perceptions of self-instructional radio (SIR) materials by learners of English in the Japanese context and factors affecting completion and non-completion of the materials. The study found that the major advantage of the materials is that the fixed on-air schedule helps learners maintain a regular learning schedule. However, the same characteristic was perceived by other...