2007 |
Word Sense Disambiguation with Spreading Activation Networks Generated from Thesauri (Paper in Conference Proceedings) Tsatsaronis, George; Varzigiannis, Michalis; Androutsopoulos, Ion Proceedings of the 20th International Joint Conference on Artificial Intelligence (IJCAI 2007), Pages: 1725-1730, 2007. (Abstract | Links | BibTeX) @inproceedings{Tsatsaronis2007,
title = {Word Sense Disambiguation with Spreading Activation Networks Generated from Thesauri},
author = {George Tsatsaronis and Michalis Varzigiannis and Ion Androutsopoulos},
url = {http://www.aueb.gr/users/ion/docs/ijcai2007_wsd_paper.pdf},
year = {2007},
date = {2007-01-01},
booktitle = {Proceedings of the 20th International Joint Conference on Artificial Intelligence (IJCAI 2007)},
pages = {1725-1730},
abstract = {Most word sense disambiguation (WSD) methods require large quantities of manually annotated training data and/or do not exploit fully the semantic relations of thesauri. We propose a new unsupervised WSD algorithm, which is based on generating Spreading Activation Networks (SANs) from the senses of a thesaurus and the relations between them. A new method of assigning weights to the networks’ links is also proposed. Experiments show that the algorithm outperforms previous unsupervised approaches to WSD.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
Most word sense disambiguation (WSD) methods require large quantities of manually annotated training data and/or do not exploit fully the semantic relations of thesauri. We propose a new unsupervised WSD algorithm, which is based on generating Spreading Activation Networks (SANs) from the senses of a thesaurus and the relations between them. A new method of assigning weights to the networks’ links is also proposed. Experiments show that the algorithm outperforms previous unsupervised approaches to WSD.
|
Η ψηφιακή επιμέλεια των πολιτισμικών συλλογών ως πεδίο γνώσης και πρακτικής (Paper in Conference Proceedings) Dallas, Costis Πρακτικά Ημερίδας της Ελληνικής Ομπσπονδίας Σωματείων Φίλων των Μουσείων (ΕΟΣΦΙΜ), 2007. (Links | BibTeX) @inproceedings{Dallas2007,
title = {Η ψηφιακή επιμέλεια των πολιτισμικών συλλογών ως πεδίο γνώσης και πρακτικής},
author = {Costis Dallas},
url = {http://www.heritage-museums.com/gr/attachments/article/359/praktika.pdf},
year = {2007},
date = {2007-01-01},
booktitle = {Πρακτικά Ημερίδας της Ελληνικής Ομπσπονδίας Σωματείων Φίλων των Μουσείων (ΕΟΣΦΙΜ)},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
|
2006 |
A Greek Named-Entity Recognizer That Uses Support Vector Machines and Active Learning (Paper in Conference Proceedings) Lucarreli, Giorgos; Androutsopoulos, Ion Advances in Artificial Intelligence. 4th Hellenic Conference on Artificial Intelligence (SETN 2006). Proceedings, Volume: 3955 of the series Lecture Notes in Computer Science Pages: 203-213, 2006. (Abstract | Links | BibTeX) @inproceedings{Lucarreli2006,
title = {A Greek Named-Entity Recognizer That Uses Support Vector Machines and Active Learning},
author = {Giorgos Lucarreli and Ion Androutsopoulos},
url = {http://www.aueb.gr/users/ion/docs/setn2006_paper.pdf},
year = {2006},
date = {2006-01-02},
booktitle = {Advances in Artificial Intelligence. 4th Hellenic Conference on Artificial Intelligence (SETN 2006). Proceedings},
volume = {3955},
pages = {203-213},
series = {Lecture Notes in Computer Science},
abstract = {We present a named-entity recognizer forGreek person names and temporal expressions. For temporal expressions, it relies on semiautomatically produced patterns. For person names, it employs two Support Vector Machines, that scan the input text in two passes, and active learning, which reduces the human annotation effort during training.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
We present a named-entity recognizer forGreek person names and temporal expressions. For temporal expressions, it relies on semiautomatically produced patterns. For person names, it employs two Support Vector Machines, that scan the input text in two passes, and active learning, which reduces the human annotation effort during training.
|
2005 |
A Practically Unsupervised Learning Method to Identify Single-Snippet Answers to Definition Questions on theWeb (Paper in Conference Proceedings) Androutsopoulos, Ion; Galanis, Dimitris Proceedings of the conference on Human Language Technology and Empirical Methods in Natural Language Processing, of the series HLT '05 Pages: 323-330, Association for Computational Linguistics, 2005. (Abstract | Links | BibTeX) @inproceedings{Androutsopoulos2005,
title = {A Practically Unsupervised Learning Method to Identify Single-Snippet Answers to Definition Questions on theWeb},
author = {Ion Androutsopoulos and Dimitris Galanis},
url = {http://www.aueb.gr/users/ion/docs/hltemnlp2005_paper.pdf},
year = {2005},
date = {2005-01-03},
booktitle = {Proceedings of the conference on Human Language Technology and Empirical Methods in Natural Language Processing},
pages = {323-330},
publisher = {Association for Computational Linguistics},
series = {HLT '05},
abstract = {We present a practically unsupervised learning method to produce single-snippet answers to definition questions in question answering systems that supplement Web search engines. The method exploits on-line encyclopedias and dictionaries to generate automatically an arbitrarily large number of positive and negative definition examples, which are then used to train an SVM to separate the two classes. We show experimentally that the proposed method is viable, that it outperforms the alternative of training the system on questions and news articles from TREC, and that it helps the search engine handle definition questions significantly better.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
We present a practically unsupervised learning method to produce single-snippet answers to definition questions in question answering systems that supplement Web search engines. The method exploits on-line encyclopedias and dictionaries to generate automatically an arbitrarily large number of positive and negative definition examples, which are then used to train an SVM to separate the two classes. We show experimentally that the proposed method is viable, that it outperforms the alternative of training the system on questions and news articles from TREC, and that it helps the search engine handle definition questions significantly better.
|
2004 |
Learning to Identify Single-Snippet Answers to Definition Questions (Paper in Conference Proceedings) Miliaraki, Spyridoula; Androutsopoulos, Ion Proceedings of the 20th International Conference on Computational Linguistics (COLING 2004), Pages: 1360-1366, 2004. (Abstract | Links | BibTeX) @inproceedings{Miliaraki2004,
title = {Learning to Identify Single-Snippet Answers to Definition Questions},
author = {Spyridoula Miliaraki and Ion Androutsopoulos},
url = {http://www.aueb.gr/users/ion/docs/coling04_definition_questions.pdf},
year = {2004},
date = {2004-01-01},
booktitle = {Proceedings of the 20th International Conference on Computational Linguistics (COLING 2004)},
pages = {1360-1366},
abstract = {We present a learning-based method to identify single-snippet answers to definition questions in question answering systems for document collections. Our method combines and extends two previous techniques that were based mostly on manually crafted lexical patterns and WordNet hypernyms. We train a Support Vector Machine (SVM) on vectors comprising the verdicts or attributes of the previous techniques, and additional phrasal attributes that we acquire automatically. The SVM is then used to identify and rank single 250-character snippets that contain answers to definition questions. Experimental results indicate that our method clearly outperforms the techniques it builds upon.},
keywords = {},
pubstate = {published},
tppubtype = {inproceedings}
}
We present a learning-based method to identify single-snippet answers to definition questions in question answering systems for document collections. Our method combines and extends two previous techniques that were based mostly on manually crafted lexical patterns and WordNet hypernyms. We train a Support Vector Machine (SVM) on vectors comprising the verdicts or attributes of the previous techniques, and additional phrasal attributes that we acquire automatically. The SVM is then used to identify and rank single 250-character snippets that contain answers to definition questions. Experimental results indicate that our method clearly outperforms the techniques it builds upon.
|