Wyniki wyszukiwania - Biblioteka Nauki

1

Evaluating lexicographer controlled semi-automatic word sense disambiguation method in a large scale experiment

100%

Broda B. , Piasecki M.

Control and Cybernetics

|

2011

|

tom Vol. 40, no 2

419-436

EN

Word Sense Disambiguation in text remains a difficult problem as the best supervised methods require laborious and costly manual preparation of training data. On the other hand, the unsupervised methods yield significantly lower precision and produce results that are not satisfying for many applications. Recently, an algorithm based on weakly-supervised learning for WSD called Lexicographer-Controlled Semi-automatic Sense Disambiguation (LexCSD) was proposed. The method is based on clustering of text snippets including words in focus. For each cluster we find a core, which is labelled with a word sense by a human, and is used to produce a classifier. Classifiers, constructed for each word separately, are applied to text. The goal of this work is to evaluate LexCSD trained on large volume of untagged text. A comparison showed that the approach is better than most frequent sense baseline in most cases.

2

On German verb sense disambiguation: A three-part approach based on linking a sense inventory (GermaNet) to a corpus through annotation (TGVCorp) and using the corpus to train a VSD classifier (TTvSense)

100%

Mattern D. , Hemati W. , Lücking A. , Mehler A.

Journal of Language Modelling

|

2024

|

tom Vol. 12, No. 1

155--212

EN

We develop a three-part approach to Verb Sense Disambiguation (VSD) in German. After considering a set of lexical resources and corpora, we arrive at a statistically motivated selection of a subset of verbs and their senses from GermaNet. This sub-inventory is then used to disambiguate the occurrences of the corresponding verbs in a corpus resulting from the union of TüBa-D/Z, Salsa, and E-VALBU. The corpus annotated in this way is called TGVCorp. It is used in the third part of the paper for training a classifier for VSD and for its comparative evaluation with a state-of-the-art approach in this research area, namely EWISER. Our simple classifier outperforms the transformer-based approach on the same data in both accuracy and speed in German but not in English and we discuss possible reasons.

3

Word Sense Disambiguation by Machine Learning Approac : A Short Survey

100%

Tatar D.

|

2005

|

tom Vol. 64, nr 1-4

433-442

EN

There is a renewed interest in word sense disambiguation (WSD) as it contributes to various applications in natural language processing. In this paper we survey vector-based methods for WSD in machine learning. All the methods are corpus-based and use definition of context in the sense introduced by S. Marcus [11].

4

Adjective Sense Disambiguation at the Border Between Unsupervised and Knowledge-Based Techniques

88%

Hristea F. , Popescu M.

|

tom Vol. 91, nr 3-4

547-562

EN

The present paper extends a new word sense disambiguation method [9] to the case of adjectives. The method lies at the border between unsupervised and knowledge-based techniques. It performs unsupervised word sense disambiguation based on an underlying Näive Bayes model, while using WordNet as knowledge source for feature selection. The proposed extension of the disambiguation method makes ample use of the WordNet semantic relations that are typical of adjectives. Its performance is compared to that of previous approaches that rely on completely different feature sets. Test results show that feature selection using a knowledge source of type WordNet is more effective in the disambiguation of adjective senses than local type features (like part-of-speech tags) are.

5

Clustering for clarity: improvingword sense disambiguation throughmultilevel analysis

75%

Dubey S. K. , Kohli N.

Computer Science

|

2024

|

tom T. 25 (2)

1--24

EN

Existing Word Sense Disambiguation (WSD) techniques have limits in exploring word-context relationships since they only deal with the effective use ofword embedding, lexical-based information via WordNet, or the precision ofclustering algorithms. In order to overcome this limitation, the study suggestsa unique hybrid methodology that makes use of context embedding based on center-embedding in order to capture semantic subtleties and utilizing with atwo-level K-means clustering algorithm. Such generated context embedding,producing centroids that serve as representative points for semantic information inside clusters. Additionally, go with such captured cluster- centres in thenested levels of clustering process, locate groups of linked context words andcategorize them according to their word meanings that effectively manage polysemy/homonymy as well as detect minute differences in meaning. Our proposedapproach offers a substantial improvement over traditional WSD methods byharnessing the power of center-embedding in context representation, enhancingthe precision of clustering and ultimately overcoming existing limitations incontext-based sense disambiguation.

6

A method to integrate word sense disambiguation and translation memory for english to hindi machine translation system

63%

Rawat S.

Computer Assisted Methods in Engineering and Science

|

2022

|

tom Vol. 29, no. 1-2 spec.

125--144

EN

Word sense disambiguation deals with deciding the word’s precise meaning in a certainspecific context. One of the major problems in natural language processing is lexical-semantic ambiguity, where a word has more than one meaning. Disambiguating the senseof polysemous words is the most important task in machine translation. This researchwork aims to design and implement English to Hindi machine translation. The designmethodology addresses improving the speed and accuracy of the machine translation process. The algorithm and modules designed in this research work have been deployed on theHadoop infrastructure, and test cases are designed to check the feasibility and reliabilityof this process. The research work presented describes the methodologies to reduce datatransmission by adding a translation memory component to the framework. The speed ofexecution is increased by replacing the modules in the machine translation process withlightweight modules, which reduces infrastructure and execution time.