From the beginning, a primary goal of the Cyc project has been to build a large knowledge base containing a store of

This paper compares a range of methods for classifying words based on linguistic diagnostics, focusing on the task of learning countabilities for English nouns. We propose two basic approaches to feature representation: distribution-based representation, which simply looks at the distribution of features in the corpus data, and agreement-based representation which analyses the level of tokenwise agreement between multiple preprocessor systems. We additionally compare a single multiclass classifier architecture with a suite of binary classifiers, and combine analyses from multiple preprocessors. Finally, we present and evaluate a feature selection method.

This thesis mainly concerns itself with the issue of semantic similarity and computational applications of it. Semantic similarity has for a long time been a subject of intense scholarship in the fields of Artificial Intelligence, Psychol-ogy and Cognitive Science. Computational models trying to imitate aspects of this cognitive ability date back to Quillian and his spreading activation algorithm. In the research presented here we propose a new computational approach to the assessment of semantic similarity between concepts. This work builds upon previous information theoretic attempts, from researchers interested in natural language processing, to mimic this ill-defined human aptitude. Word-Net, an ontological lexical knowledge base, is the only source of knowledge exploited in the proposed algorithm. Hence, this is where our work differs

We examine the hypothesis that noun countability is consistent for a given word semantics by way of a series of experiments involving EuroWordNet and the English and Dutch languages. The basic method involves determining a default set of countabilities for each EuroWordNet synset based on countability-mapped words in that synset, and testing the match between these countabilities and those of held-out words. As EuroWordNet provides crosslingual synset correspondences between Dutch and English, we are able to evaluate the method both monolingually for Dutch and English, and crosslingually between the two languages. We found that Dutch and English countabilities align as well cross-lingually as they do monolingually.

Use cases are commonly used to represent customers’ requirements during systems development. In a large software development environment, finding a relevant use case from a large use case library created in the past or related projects is a complex, error-prone and expensive task. Based on the semantic web approach, we propose an ontological methodology to support this task. We use an ontology to augment use cases with semantic information. This ontology is derived from ResearchCyc ontology. We also propose the augmentation of queries used to retrieve use cases with this ontology. We present this approach to better capture, reuse and query use cases 1.

We present an approach for learning criteria for part-of-speech classification by induction over the lexicon contained within the Cyc knowledge base.

We present an approach to learning criteria for classifying the parts-of-speech used in lexical mappings. This will further automate our knowledge acquisition system for nontechnical users. The criteria for the speech parts are based on the types of the denoted terms along with morphological clues and are induced from the lexical mappings contained in the Cyc knowledge base. With over 30 speech parts to choose from, the classifier achieves good results (73.3% correct). Accurate results (90.5%) are achieved in the special case of the mass-count distinction for nouns. Comparable results are obtained using OpenCyc(69.3% general and 87.5% mass-count). Similar results are achieved with criteria mapped into WordNet (67.0% general and 86.3% mass-count).

Countability of English nouns is important in various natural language processing tasks. It especially plays an important role in machine translation since it determines the range of possible determiners. This paper proposes a method for reinforcing countability prediction by introducing a novel concept called one countability per discourse. It claims that when a noun appears more than once in a discourse, they will all share the same countability in the discourse. The basic idea of the proposed method is that mispredictions can be correctly overridden using efficiently the one countability per discourse property. Experiments show that the proposed method successfully reinforces countability prediction and outperforms other methods used for comparison. 1