Title: Simulation of language competition by bilingual agents

Authors: Rakesh Kumar; ShivKishan Dubey

Addresses: Department of Computer Science and Engineering, National Institute of Technical Teachers Training and Research, Sector-26, Chandigarh 160019, India ' Department of Computer Science and Engineering, Dr. Ambedkar Institute of Technology for Handicapped, Awadhpuri, Kanpur 208024, India

Abstract: In the language transmission, a person shares different ways of own lexicons to others and each one is able to reproduce new lexicons through the old ones. Our objective is to explore how lexicons can be modified in synchronic or diachronic scenarios during the interaction of two different languages (L1 or L2) community agents. In this research, a new model has been proposed within language competition game strategy using bilingual agents. The outcome of shared lexicons in the synchronic scenario is that massive codes of second-language are affected and propagated into future generation successfully. Initially, we take matured agents who are stimulated by iterated learning model (ILM) and equipped with own grammar rules. Using Levenshtein Distance, these agents modify own old lexicons and generate new ones for the next generation. This game strategy is restricted to synchronic only and found that second language always wins this game.

Keywords: NLP; natural language processing; iterated learning model; language acquisition; bilingualism; exposure ratio.

DOI: 10.1504/IJKL.2018.095979

International Journal of Knowledge and Learning, 2018 Vol.12 No.4, pp.362 - 378

Received: 29 Jul 2017
Accepted: 07 Aug 2018
Published online: 05 Nov 2018 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article