Steels, L., (1997), Language Learning and Language Contact, in Daelemans, W. (ed.), (1997), Proceedings of the workshop on Empirical Approaches to Language Acquisition, Prague.  Home/Search   Document Details and Download   Summary   Related Articles   Check

....on the experimentation with Artificial Languages (ALanguages) A Languages use computational approaches to the problem of explaining the origin and evolution of language. In the field of the origins of language Luc Steels and the Origins of Language Group investigate linguistic phenomena [1, 2]. Mike Oliphant, Simon Kirby and Jim Hurford have applied mathematical and computational modeling techniques to traditional issues in the evolution of communication and language [3 7] Ezequiel di Paolo has been studying the evolution of communication, mainly through the processes of action and ....

Steels, L., (1997), Language Learning and Language Contact, in Daelemans, W. (ed.), (1997), Proceedings of the workshop on Empirical Approaches to Language Acquisition, Prague.

.... have addressed the problem essentially from an evolutionary perspective, 1) either through theoretical models based on biological studies of brain evolution (e.g. 16, 46] or on sociological studies of primates and human societies (e.g. 2, 18] 2) or through computer simulations (e.g. [28, 34, 45, 43, 51]) For these studies, the symbol grounding problem 2 [22] is solved, once the necessary cognitive abilities have evolved. However, few of these studies consider the influence of behavioural and social factors on the development of communication, exceptions are [18, 45] A common trend among the ....

.... or strongly simplifying agent environment dynamics (e.g. assuming complete or global information of the world and other agents) The cognitive functions involved in grounding of communication have been simulated e.g. as the input output of an artificial neural network [28] as a matching process [51] or as a probability function [43] The agents in these simulations are disembodied, they do not occupy a physical space (they have no body, no sensors or actuators, and generally occupy not more than a single point in the space) and the result of their actions is atemporal (an action and its ....

[Article contains additional citation context not shown here]

Steels, L., (1997) `Language Learning and Language Contact', In: Daelemans, W. (ed.). Proceedings of the workshop on Empirical Approaches to Language Aquisition. Prague. 47

....2. Many results have already been worked out, in particular by coupling it with other language games ( 5] Yet, most of them concentrated on the language itself and did not consider the possibility of having a priori extralinguistic structure in the agent society. The first step was done in [4] where the idea of spatialization was introduced : the agent society is given a fixed topological structure (that can be based on a geographical or social distance for instance) that determines with who agents may speak, as we will see in section 2. We will present new results in section 3, based ....

....IS and SC. When parameters are not precised, it means that they have default value : 100 agents randomly located in 2 D space, 4 meanings, loc = 8, w a = 0:1, w c = 0:1, IS = 0:1 and SC = 0:1. The first results yielded by those experiments are that like in the non spatialized naming game (NSNG) [4]) success reaches more than 85 percent after 1500 games, and this at the same speed. But unlike NSNG, where one form rapidly dominates for a given meaning, several stable forms appear here and go to an equilibrium where each one keeps its percentage of users : Figure 1 shows formspread for one of ....

L. Steels, 1997, Language Learning and Language Contact. In : W. Daelemans, A. Van den Bosh and A. Weijters (eds.), Worshop Notes of the ECML/MLnet Familiarization Workshop on Empirical Learning of Natural Language Processing Tasks. Prague. pp 11-24.

....situations, an outside observer can observe indications that communication or coordination takes place. We use the term measures for these observable indications. Measures are defined quantitatively and can thus be determined exactly. For examples of measures used in this research programme, see [Steels, 1997], Kaplan, 1998] Measures need to be defined specific to each experiment. The difficulties with defining communication result from the wish to find general operational criteria that determine the presence of communication in all its appearances. I suggest that these problems can be circumvented ....

Steels, L. (1997). Language learning and language contact. In Daelemans, W., van den Bosch, A., and Weijters, A., editors, Workshop Notes of the ECML/MLnet Familiarization Workshop on Empirical Learning of Natural Language Processing Tasks, pages 11--24.

....and can thus be determined exactly. Using these measures, researchers can investigate communication in many different types of experiments, even though a general definition of communication is not available. Within the origins of language programme, measures are used extensively, see e.g. [17], 8] furthermore, see [11] Another way to test whether agents communicate is to replicate the supposedly communicative behavior. If this induces the predicted effect, it indicates that the behavior under consideration was indeed used for communication. This method has been demonstrated ....

L. Steels. Language learning and language contact. In W. Daelemans, A. van den Bosch, and A. Weijters, editors, Workshop Notes of the ECML/MLnet Familiarization Workshop on Empirical Learning of Natural Language Processing Tasks, pages 11--24, 1997.

....indications that communication takes place. We will define criteria, which we call measures, that indicate whether communication is taking place or not. These measures are defined quantitatively and can thus be determined exactly. For examples of measures used in this research programme, see [Steels, 1997], Kaplan, 1997] We define communication as follows: Definition1. Communication is the successful transmission of meaning from one agent to one or more other agents. In the above, it has become clear that meaning can sometimes refer to implicit information, and at other times to explicit ....

Steels, L. (1997). Language learning and language contact. In Daelemans, W., van den Bosch, A., and Weijters, A., editors, Workshop Notes of the ECML/MLnet Familiarization Workshop on Empirical Learning of Natural Language Processing Tasks, pages 11--24.

No context found.

Steels, L. (1997b) Language Learning and Language Contact. In: Daelemans, W. (ed.) Empirical approaches to Language Learning. ECML Prague.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC