In this paper, we present a natural generalization of k-gram models for tree stochastic languages based on the k-testable class. In this class of models, frequencies are estimated for a probabilistic regular tree grammar wich is bottom-up deterministic. One of the advantages of this approach is that the model can be updated in an incremental fashion. This method is an alternative to costly learning algorithms (as inside-outside-based methods) or algorithms that require larger samples (as many state merging/splitting methods).

To learn a language, the learners must first learn its words, the essential building blocks for utterances. The difficulty in learning words lies in the unavailability of explicit word boundaries in speech input. The learners have to infer lexical items with some innately endowed learning mechanism(s) for regularity detection- regularities in the speech normally indicate word patterns. With respect to Zipf's least-effort principle and Chomsky's thoughts on the minimality of grammar for human language, we hypothesise a cognitive mechanism underlying language learning that seeks for the least-effort representation for input data. Accordingly, lexical learning is to infer the minimal-cost representation for the input under the constraint of permissible representation for lexical items. The main theme of this thesis is to examine how far this learning mechanism can go in unsupervised lexical learning from real language data without any pre-defined (e.g., prosodic and phonotactic) cues, but entirely resting on statistical induction of structural patterns for the most economic representation for the data. We first review

In this dissertation, we address the problem of recognizing human interactions with objects from video. Methods for recognizing these activities using human motion and information about objects are developed for practical, real-time systems. We introduce a framework, called ObjectSpaces, that sorts, stores, and manages data acquired using low-level vision techniques into intuitive classes. Our framework decomposes the recognition process into layers, i.e., a low-level layer for routine hand and object tracking and a high-level layer for domain-specific representation of activities. Segmenting recognition tasks and information in this way encourages model reuse and provides the flexibility to use a single framework in a variety of domains. We present several ways of...

This paper describes three preliminary experiments in adding new language knowledge to the recognizer BeRP:

Abstract not found

We consider the problem of computing the Kullback-Leibler distance, also called the relative entropy, between a probabilistic context-free grammar and a probabilistic finite automaton. We show that there is a closed-form (analytical) solution for one part of the Kullback-Leibler distance, viz. the cross-entropy. We discuss several applications of the result to the problem of distributional approximation of probabilistic context-free grammars by means of probabilistic finite automata. 1

Probabilistic k-testable models (usually known as k-gram models in the case of strings) can be easily identified from samples and allow for smoothing techniques to deal with unseen events during pattern classification. In this paper, we introduce the family of stochastic k-testable tree languages and describe how these models can approximate any stochastic rational tree language. The model is applied to the task of learning a probabilistic k-testable model from a sample of parsed sentences. In particular, a parser for a natural language grammar that incorporates smoothing is shown.

. In this paper, we present a natural generalization of k-gram models for tree stochastic languages based on the k-testable class. In this class of models, frequencies are estimated for a probabilistic regular tree grammar wich is bottom-up deterministic. One of the advantages of this approach is that the model can be updated in an incremental fashion. This method is an alternative to costly learning algorithms (as inside-outside-based methods) or algorithms that require larger samples (as many state merging/splitting methods). 1 Introduction Stochastic models based on k-grams have been widely used in natural language modeling [BPd + 92, NEK95], speech recognition [Jel98] and data compression [Rub76]. Indeed, any stochastic model can be used to predict the next symbol in a sequence and, therefore, they are a suitable component in arithmetic data compression [CT91] algorithms. In classication problems, the need of stochastic models often arises when the Bayes' decision rule ...

ion. In the mean while, we refer the reader to [Jelinek & Mercer 1980] for Deleted Interpolation, to [Good 1953] and [Katz 1987] for Good-Turing reestimation, and to [Samuelsson 1996] for Successive Abstraction. 58 CHAPTER 2. APPLIED STATISTICS Chapter 3 Basic Corpus Linguistics 3.1 Empirical Evaluation Chinchor et al. write in Computational Linguistics 19(3): "One of the common problems with evaluations is that the statistical significance of the results is unknown", [Chinchor et al 1993], p. 409. Empirical evaluation of natural language processing systems in general is very young. The attempt to put evaluation of the efficiency of speech recognition, and natural language processing systems on solid grounds started at the end 80ies in the US within the (D)ARPA 1 speech and natural language workshops as well as the (D)ARPA message understanding conferences. First evaluation settings were so called blackbox evaluations, i.e. the systems under evaluation are tested as a whole, n...

This dissertation is dedicated to my parents, both of whom earned advanced degrees late in life and especially to my wife, Corinne, whose love and support have always been greater than I deserve. ii Contents Tables................................................................................................................ v Figures............................................................................................................. vi