Two graph-based algorithms for state-of-the-art WSD

Agirre, Eneko; Martínez, David; Lacalle, Oier López de; Soroa, Aitor

doi:10.3115/1610075.1610157

Cited by 57 publications

(64 citation statements)

References 13 publications

(26 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Graph-based clustering for WSI has a long history and many different variations (Lin et al, 1998;Pantel and Lin, 2002; Dorow and Widdows, 2003;Véronis, 2004;Agirre et al, 2006; Biemann, 2006;Navigli and Crisafulli, 2010; Hope and Keller, 2013; Di Marco and Navigli, 2013;Mitra et al, 2014;Pelevina et al, 2016). In general, the method is to first retrieve words similar or related to each target word as nodes, measure the similarity/relatedness between the words to form an ego graph/network, and either group the nodes by graph clustering or find hubs or representative nodes in the graph using HyperLex (Véronis, 2004) or PageRank (Agirre et al, 2006).…”

Section: Clustering Related Wordsmentioning

confidence: 99%

“…In general, the method is to first retrieve words similar or related to each target word as nodes, measure the similarity/relatedness between the words to form an ego graph/network, and either group the nodes by graph clustering or find hubs or representative nodes in the graph using HyperLex (Véronis, 2004) or PageRank (Agirre et al, 2006).…”

Section: Clustering Related Wordsmentioning

confidence: 99%

“…The problem is sometimes also called unsupervised word sense disambiguation (Agirre et al, 2006;Pelevina et al, 2016). An effective WSI has wide applications.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Proceedings of the Twelfth Workshop on Graph-Based Methods for Natural Language Processing (TextGraphs-12)

2018

View full text Add to dashboard Cite

The workshops in the TextGraphs series have published and promoted the synergy between the field of Graph Theory (GT) and Natural Language Processing (NLP) for over a decade. The target audience of our workshop comprises of researchers working on problems related to either Graph Theory or graphbased algorithms applied to Natural Language Processing, social media, and the Semantic Web.TextGraphs addresses a broad spectrum of research areas within NLP. This is because, besides traditional NLP applications like parsing, word sense disambiguation, semantic role labeling, and information extraction, graph-based solutions also target web-scale applications like information propagation in social networks, rumor proliferation, e-reputation, language dynamics learning, and future events prediction. Following this tradition, this year's TextGraphs also presents research from diverse topics such as lexical and computational semantics, text clustering and classification, and text compresion and summarization, to name a few.The selection process was competitive -we received 17 submissions (9 long and 8 short submissions) and accepted for presentation 8 of them (5 long and 3 short papers), resulting in the overall acceptance rate of 47%.We are pleased to have two excellent invited speakers for this year's event. We thank Gabor Melli and Maximilian Nickel for their enthusiastic acceptance of our invitation. Finally, we are thankful to the members of the program committee for their valuable and high quality reviews. All submissions have benefited from their expert feedback. Their timely contribution was the basis for accepting an excellent list of papers and making the twelfth edition of TextGraphs a success. AbstractWe introduce a machine learning approach for the identification of "white spaces" in scientific knowledge. Our approach addresses this task as link prediction over a graph that contains over 2M influence statements such as "CTCF activates FOXA1", which were automatically extracted using open-domain machine reading. We model this prediction task using graph-based features extracted from the above influence graph, as well as from a citation graph that captures scientific communities. We evaluated the proposed approach through backtesting. Although the data is heavily unbalanced (50 times more negative examples than positives), our approach predicts which influence links will be discovered in the "near future" with a F1 score of 27 points, and a mean average precision of 68%. IntroductionThe amount of scientific knowledge that is publicly available has increased dramatically in the past few years. For example, PubMed, a search engine of biomedical publications, 1 now indexes over 25 million papers, 17 million of which were published between 1990 and the present. This information overload yields two critical problems. First, this exceeds the human capacity to aggregate and interpret the fragments of knowledge published in these papers, which may result in existing solutions to critical problems being overlooked. Swan...

show abstract

Section: Clustering Related Wordsmentioning

confidence: 99%

Section: Clustering Related Wordsmentioning

confidence: 99%

See 1 more Smart Citation

Proceedings of the Twelfth Workshop on Graph-Based Methods for Natural Language Processing (TextGraphs-12)

2018

View full text Add to dashboard Cite

show abstract

“…Then vectors are clustered and the resulting clusters represent the word senses. Recently some works have developed graph-based methods to achieve WSI [13,29,30]. Typically these works select contexts of a given ambiguous word w and assign every word appearing in these contexts to a node of the graph.…”

Section: A Descriptionmentioning

confidence: 99%

Disentangling categorical relationships through a graph of co-occurrences

et al. 2011

View full text Add to dashboard Cite

The mesoscopic structure of complex networks has proven a powerful level of description to understand the linchpins of the system represented by the network. Nevertheless, the mapping of a series of relationships between elements, in terms of a graph, is sometimes not straightforward. Given that all the information we would extract using complex network tools depend on this initial graph, it is mandatory to preprocess the data to build it on in the most accurate manner. Here we propose a procedure to build a network, attending only to statistically significant relations between constituents. We use a paradigmatic example of word associations to show the development of our approach. Analyzing the modular structure of the obtained network we are able to disentangle categorical relations, disambiguating words with success that is comparable to the best algorithms designed to the same end.

show abstract

“…In a graph of this type, the vertices correspond to the words appearing in the contexts of the target words and the edges represent their relations. These relations may be grammatical [12] or they may be cooccurrences of the words in fixed contexts [13,14]. The senses of the target words are discovered by partitioning the co-occurrence graph using clustering techniques, or by using a PageRank algorithm.…”

Section: Inducing Word Senses On a Per-word Basismentioning

confidence: 99%

A Quantitative Evaluation of Global Word Sense Induction

Apidianaki

Cruys

2011

Computational Linguistics and Intelligent Text Processing

View full text Add to dashboard Cite

Abstract. Word sense induction (WSI) is the task aimed at automatically identifying the senses of words in texts, without the need for handcrafted resources or annotated data. Up till now, most WSI algorithms extract the different senses of a word 'locally' on a per-word basis, i.e. the different senses for each word are determined separately. In this paper, we compare the performance of such algorithms to an algorithm that uses a 'global' approach, i.e. the different senses of a particular word are determined by comparing them to, and demarcating them from, the senses of other words in a full-blown word space model. We adopt the evaluation framework proposed in the SemEval-2010 Word Sense Induction & Disambiguation task. All systems that participated in this task use a local scheme for determining the different senses of a word. We compare their results to the ones obtained by the global approach, and discuss the advantages and weaknesses of both approaches.

show abstract

Two graph-based algorithms for state-of-the-art WSD

Cited by 57 publications

References 13 publications

Proceedings of the Twelfth Workshop on Graph-Based Methods for Natural Language Processing (TextGraphs-12)

Proceedings of the Twelfth Workshop on Graph-Based Methods for Natural Language Processing (TextGraphs-12)

Disentangling categorical relationships through a graph of co-occurrences

A Quantitative Evaluation of Global Word Sense Induction

Contact Info

Product

Resources

About