Knowledge level learning and the power law: A soar model of skill acquisition in scheduling

Zusammenfassung. Das hier vorgestellte Modell lernt graduell, Planungsaufgaben aus der Klasse der Maschinenbelegungsprobleme (job-shop-scheduling problems) zu lösen. Mit Hilfe des Chunking-Mechanismus von Soar wird episodisches Wissen über die Belegungsreihenfolge von Aufträgen auf Maschinen memoriert. Bei der Entwicklung des Modells wurden zahlreiche qualitative (z. B. Transfereffekte) und quantitative Befunde (z. B. Bearbeitungszeiten) aus einer früheren empirischen Untersuchung berücksichtigt. In einer Validierungsstudie wurden dieselben Aufgaben von 14 Probanden und dem Modell bearbeitet. Die Passung von Simulationsdaten und empirischen Ergebnissen fiel insgesamt gut aus. Allerdings löst das Modell die Aufgaben schneller und zeigt auch einen etwas besseren Lernverlauf als die Probanden. Das Modell liefert eine Erklärung für das Rauschen, das typischerweise bei Bearbeitungszeiten zu beobachten ist: es handelt sich um erworbenes Wissen, das mehr oder weniger gut und auch unterschiedlich häufig auf neue Situationen übertragen wird. Der Lernverlauf der Probanden entspricht nur für aggregierte Daten einer Potenzfunktion (power law). Der vorgestellte Mechanismus zeigt, wie ein symbolisches Modell der Informationsverarbeitung graduelle Verhaltensänderungen generiert und wie der offensichtliche Erwerb allgemeiner Prozeduren ohne explizites Lernen von deklarativen Regeln erfolgen kann. Es wird nahegelegt, daß es sich hier um die Modellierung einer Form impliziten Lernens handelt. Summary. The model presented here gradually learns how to perform a job-shop scheduling task. It uses Soar's chunking mechanism to acquire episodic memories about the order to schedule jobs. The model was based on many qualitative (e.g., transfer effects) and quantitative (e.g., solution time) regularities found in previously collected data. The model was tested with new data where scheduling tasks were given to the model and to 14 subjects. The model generally fit these data with the restrictions that the model performs the task (in simulated time) faster than the subjects, and its performance improves somewhat more quickly than the subjects' performance. The model provides an explanation of the noise typically found in problem solving times - it is the result of learning actual pieces of knowledge that transfer more or less to new situations but rarely by an average amount. Only when the data are averaged (i.e., over subjects) does the smooth power law appear. This mechanism demonstrates how symbolic models can exhibit a gradual change in behavior and how the apparent acquisition of general procedures can be performed without resorting to explicit declarative rule generation. We suggest that this may represent a type of implicit learning.     

The puzzle of language evolution

Linguistics must again concentrate on the evolutionary nature of language, so that language models are more realistic with respect to human natural languages and have a greater explanatory force. Multi-agent systems are proposed as a possible route to develop such evolutionary models and an example is given of a concrete experiment in the origins and evolution of word-meaning based on a multi-agent approach.     

Interactive learning in the paradise fish (Macropodus opercularis): an ethological interpretation of the second-order conditioning paradigm

This study was aimed at the examination of ‘mental construction’ in paradise fish by interactive learning, which is suggested as an alternative hypothesis for backward or second-order conditioning. Avoidance of goldfish was established in paradise fish by presenting a harmless goldfish (a novel stimulus) after an aversive event (mild electric shocks) in the dark compartment of a shuttle tank. It was found that this avoidance depended on context pre-exposure. Experiment 1 was designed to study the effect of mild shocks on shuttling activity in a familiar context. Experiment 2 aimed at establishing fear-conditioning to the goldfish in a higher-order conditioning paradigm. In the course of training, unpaired stimuli were presented in the dark compartment of the shuttle tank in such a manner that the presentation of mild shocks (20 mA) preceded the encounter with a harmless fish (goldfish). Experiment 3 demonstrated the role of context pre-exposure in interactive learning. Results indicate that (1) while 60 mA shocks resulted in avoidance of the dark compartment, the 20 mA mild shocks affected exploratory behaviour; (2) after pre-exposure to the training environment, paradise fish avoided the dark compartment containing goldfish provided that subjects had previously encountered mild, explorative shocks there; (3) this conditioned fear, a ‘mental construction’ of a potential danger, was proved to be transferable to another context and was consequently aimed specifically at the goldfish, a living being, rather than the training context; (4) the pre-exposure to the shuttle tank was an important part of this training procedure, that is, only subjects habituated to the shuttle tank avoided the goldfish. Results are discussed in the framework of the Interactive Learning Hypothesis, which has been developed as an ethological approach to a higher-order conditioning paradigm. Received: 15 January 1999 / Accepted after revision: 18 August 1999     

Evidence for abstract, schematic knowledge of three spatial diagram representations

Spatial diagram representations such as hierarchies, matrices, and networks are important tools for thinking. Our data suggest that college students possess abstract schemas for these representations that include at least rudimentary information about their applicability conditions. In Experiment 1, subjects were better able to select the appropriate spatial diagram representation for a problem when cued to use general category information in memory about those representations than when cued to use specific example problems given during the experiment. The results of Experiment 2 showed that the superior performance in the general category condition was not based on a comparison of the test problems with examples in memory. The results of Experiment 3 showed that the superior performance was not due to learning that occurred during the experiment or to transfer appropriate processing. The General Discussion section considers the nature of students' representation schemas and the question of why college students have only rudimentary schemas for common and widely applicable diagrammatic representations.     

Attributions and affective reactions of family members and course of schizophrenia.

The authors tested an attribution-affect model of schizophrenic relapse attending to the role of families' positive affect (warmth) and negative affect (criticism). Coders listened to interviews of 40 family members taken from C. E. Vaughn, K. S. Synder, S. Jones, W. B. Freeman, and I. R. Falloon (1984) and rated their attributions of controllability for the symptoms and behaviors of their relatives with schizophrenia. For family members not designated as emotionally overinvolved, perceptions that their ill relatives' symptoms and behaviors were under the patients' control were related to family members' warmth and criticism and to patients' clinical outcomes. Of the affective reactions, only criticism predicted outcome. In addition, patients' use of street drugs was related to attributions, criticism, and outcome. Together these findings suggest that families' attributions and criticism are important in understanding the relationship between family factors and course of illness.     

Does Alternating Between Masculine and Feminine Pronouns Eliminate Perceived Gender Bias in Text?

This study explores whether alternating betweenthe pronouns he and she ina text is an effective way to avoid sexist language.Participants were psychology students at a largemidwestern university and were predominately White and frommiddle-class backgrounds. Students read two versions ofan essay, one that alternated between masculine andfeminine pronouns and one that exclusively used paired, he or she-type pronouns. Readersperceived the alternating version to be biased in favorof females and lower in overall quality than the pairedversion. However, the alternating version appeared to be more effective at combating sexism,suggesting an alternating strategy may be desirable forauthors with this goal. If the author is not primarilyconcerned with increasing readers' awareness of gender issues, techniques such as pluralization or thesingular they may be moreappropriate.