1,721,006 research outputs found
"You'd Better Say Nothing Than Say Something Wrong": Analogy, Accuracy and Text-To-Speech Applications
No full textOn the pronunciation of unknown words by analogy in text-to-speech systems: an evaluation
No full text
Resolving syntactic ambiguities with lexico-semantic patterns: an analogy-based approach
No full textA system for the resolution of syntactic ambiguities is illustrated which operates on morpho-syntactically ambiguous subject-object assignments in Italian adn tries to find the most likely analysis on the basis of the evidence contained in a knowledge base of linguistic data automatically extracted from on-line resources. The system works on the basis of a set of straightforward analogy-based principles. Its performance on a substantial corpus of test data extracted from real texts is described
Inferring semantic similarity from distributional evidence: an analogy-based approach to word sense disambiguation
No full textThe paper describes an analogy-based measure of word-sense proximity grounded on distributional evidence in typical contexts, and illustrates a computational system which makes use of this measure for purposes of lexical disambiguation. Experimental results show that word sense-analogy based on contexts of use compares favourably with classical word-sense similarity defined in terms of thesaural proximity
CHUNK-IT. An Italian Shallow Parser for Robust Syntactic Annotation
No full textThis paper reports on the experience of developing and applying a shallow parsing scheme, “chunking”, to unrestricted Italian texts, with a view to the prospective definition of further, more complex levels of syntactic analysis. A text is chunked into structured units which can be identified with certainty on the basis of an empty syntactic lexicon. The chunking process stops at that level of granularity beyond which the analysis gets undecidable. We argue that a chunked syntactic representation can usefully be exploited as such for non trivial NLP applications, which do not require full text understanding such as automatic lexical acquisition and information retrieval. The first part of the paper illustrates in detail the adopted annotation scheme, by relating it to some specific issues of Italian syntactic analysis. In the second part, after giving some theoretical justification of the notion of chunking, we describe some applications of this technique of shallow parsing to robust syntactic annotation of texts
Example-based word sense disambiguation: a paradigm-driven approach
No full textThe paper describes an example-based approach to word sense disambiguation:
words in a pre-processed text corpus are automatically linked to their corresponding
senses in a machine readable dictionary (MRD) by using information automatically
extracted from the MRD. For each word sense, typical contexts of use were acquired
and structured as "paradigmatic structures" on the basis of distributional criteria.
Word sense disambiguation is modelled as a process of "paradigm extension"
grounded on the acquired paradigmatic structures. The technique, already applied
with success to a number of Natural Language Processing (NLP) applications, is
currently under extensive test for word sense disambiguation: preliminary results
look promising
Analogy-based learning and Natural Language Processing
No full textThe role and power of analogy in the acquisition and mastering of language has been largely neglected in recent linguistic literature. An explanation can mainly be found in the inherent difficulty of defining a formal setting for a rigorous evaluation of the power of analogy, which has thus been dismissed by most formal linguists as a woolly and at best unworkable notion. Nowadays, the general availability of computers with huge and cheap storage resources appears to offer an unprecedented opportunity for an algorithmic definition of analogy and for a scientific assessment of its role in Natural Language Processing applications. We discuss recent work in this area in collaboration with the "Istituto di Linguistica Computazionale" (ILC-CNR), Pisa
- …
