Learning Meaning Before Syntax

Lecture Notes in Computer Science

2011

Self Cite

Abstract. The field of Grammatical Inference was originally motivated by the problem of natural language acquisition. However, the formal models proposed within this field have left aside this linguistic motivation. In this paper, we propose to improve models and techniques used in Grammatical Inference by using ideas coming from linguistic studies. In that way, we try to give a new bio-inspiration to this field.

Section: Bio-inspired Grammatical Inferencementioning

confidence: 86%

Bio-inspired Grammatical Inference

Lecture Notes in Computer Science

2011

Self Cite

“…This paper is an extension of our previous work on semantics and corrections [3,4,5,6,7]. In [3,4], we proposed a non-probabilistic model of meaning and denotation using finite-state transducers. We gave an algorithm to learn a meaning function and proved that it finitely converges to a correct result under a specific set of assumptions about the transducer and examples used.…”

Section: Introductionmentioning

confidence: 93%

A model of language learning with semantics and meaning-preserving corrections

Angluin

Artificial Intelligence

2017

Self Cite

We present a computational model that takes into account semantics for language learning and allows us to model meaning-preserving corrections. The model is constructed with a learner and a teacher who interact in a sequence of shared situations by producing utterances intended to denote a unique object in each situation. We test our model with limited sublanguages of 10 natural languages exhibiting a variety of linguistic phenomena. The results show that learning to a high level of performance occurs after a reasonable number of interactions. Comparing the effect of a teacher who does no correction to that of a teacher who corrects whenever possible, we show that under certain conditions corrections can accelerate the rate of learning. We also define and analyze a simplified model of a probabilistic process of collecting corrections to help understand the possibilities and limitations of corrections in our setting.

“…An example of it is [2,3,4]. The purpose of this work was not to learn a CF grammar, but to investigate the effect of semantics and corrections in the process of learning to understand and speak a natural language.…”

Section: Practical Studies With Natural Language Datamentioning

confidence: 99%

Learning, Agents, and Formal Languages: Linguistic Applications of Interdisciplinary Fields

Advances in Intelligent Systems and Computing

Jiménez-López

2015

Self Cite

Abstract. This paper focuses on three areas: machine learning, agent technologies and formal language theory. Our goal is to show how the interrelation among agents, learning and formal languages can contribute to the solution of a challenging problem: the explanation of how natural language is acquired and processed. Linguistic contributions of the intersection between machine learning and formal language theory -through the field of grammatical inference-are reviewed. Agent-based formal language models as colonies, grammar systems and eco-grammar systems have been applied to different natural language issues. We review the most relevant applications of these models.