If a word class is defined on the basis of its co-occurrence with another word class in texts, then its ranking distribution may be modelled well by the negative hypergeometric probability distribution, and the interaction of the two classes can be described by the same type of model as well as with the help of some similar procedures, such as the non-linear regression. Possessive adjectives an...

This paper investigates the perplexity and word error rate performance of two different forms of class model and the respective data-driven algorithms for obtaining automatic word classifications. The computational complexity of the algorithm for the ‘conventional’ two-sided class model is found to be unsuitable for very large vocabularies ( 100k) or large numbers of classes ( 2000). A one-side...

This paper investigates the impact on French dependency parsing of lexical generalization methods beyond lemmatization and morphological analysis. A distributional thesaurus is created from a large text corpus and used for distributional clustering and WordNet automatic sense ranking. The standard approach for lexical generalization in parsing is to map a word to a single generalized class, eit...

We characterize the infinite words determined by one-way stack automata. An infinite language L determines an infinite word α if every string in L is a prefix of α. If L is regular or context-free, it is known that α must be ultimately periodic. We extend this result to the class of languages recognized by one-way nondeterministic checking stack automata (1-NCSA). We then consider stronger clas...

We investigate co-indexed groups, that is groups whose co-word problem (all words defining nontrivial elements) is an indexed language. We show that all Higman-Thompson groups and a large class of tree automorphism groups defined by finite automata are co-indexed groups. The latter class is closely related to dynamical systems and includes the Grigorchuk 2-group and the Gupta-Sidki 3-group. The...

Word Sense Disambiguation (WSD) aims at identifying the correct sense of a word in a given context. WSD is an important task for other applications as Machine Translation or Information Retrieval. For English, WSD has been widely studied, obtaining different performances. Analyzing by morphosyntactic class, Verb is the hardest class to be disambiguated. Verbs are an important class and help to ...

In this article, we propose two algorithms for determining the Nielsen-Thurston classification of a mapping class ψ on a surface S. We start with a finite generating set X for the mapping class group and a word ψ in 〈X〉. We show that if ψ represents a reducible mapping class in Mod(S) then ψ admits a canonical reduction system whose total length is exponential in the word length of ψ. We use th...

A three layer back propagation neural net is set up to study the functional dependency between the semantic class of a bisyllabic Chinese word and that of its two constituent Chinese characters. Simulations were performed using a three-layer back-propagation neural net with various combination of inputs. The inputs are (1) semantic classes of the constituent characters, (2) Entropy of the chara...

This paper discusses the development of trainable statistical models for extracting content from television and radio news broadcasts. In particular we concentrate on statistical finite state models for identifying proper names and other named entities in broadcast speech. Two models are presented: the first represents name class information as a word attribute; the second represents both word-...

In this paper, we describe and evaluate a language model using word classes automatically generated from a word clustering algorithm. Class based language models have been shown to be effective for rapid adaptation, training on small datasets, and reduced memory usage. In terms of model perplexity, prior work has shown diminished returns for class based language models constructed using very la...