Machine learning and grammar induction

Pat Langley
1987 Machine Learning  
Language is a major component of cognition, and as such, its acquisition has been a central concern of machine learning researchers. Some of the earliest AI learning work focused on this topic, and interest has continued to the present. However, progress in this area has been much slower than in most other learning tasks, undoubtedly due to the inherent complexity of natural language. Despite its complexity, the task of natural language processing can be divided into a number of well-defined
more » ... tasks, and one of these centers on syntax. This aspect of language has been studied in detail by linguists, and developmental studies have provided a variety of empirical generalizations about the stages that children traverse in their acquisition of grammar. Because our knowledge of syntax is more complete than that for other components of language, the vast majority of language-related research in machine learning has focused on the task of grammar induction.
doi:10.1007/bf00058752 fatcat:gdowydozrvdaxoh734qvvpymka