Search Results

Documents authored by Sauri, Roser

Found 2 Possible Name Variants:

Saurí, Roser

Document
DOI: 10.4230/OASIcs.LDK.2019.20
Cross-Dictionary Linking at Sense Level with a Double-Layer Classifier

Authors: Roser Saurí, Louis Mahon, Irene Russo, and Mironas Bitinis

Published in: OASIcs, Volume 70, 2nd Conference on Language, Data and Knowledge (LDK 2019)

Abstract
We present a system for linking dictionaries at the sense level, which is part of a wider programme aiming to extend current lexical resources and to create new ones by automatic means. One of the main challenges of the sense linking task is the existence of non one-to-one mappings among senses. Our system handles this issue by addressing the task as a binary classification problem using standard Machine Learning methods, where each sense pair is classified independently from the others. In addition, it implements a second, statistically-based classification layer to also model the dependence existing among sense pairs, namely, the fact that a sense in one dictionary that is already linked to a sense in the other dictionary has a lower probability of being linked to a further sense. The resulting double-layer classifier achieves global Precision and Recall scores of 0.91 and 0.80, respectively.
Cite as

Roser Saurí, Louis Mahon, Irene Russo, and Mironas Bitinis. Cross-Dictionary Linking at Sense Level with a Double-Layer Classifier. In 2nd Conference on Language, Data and Knowledge (LDK 2019). Open Access Series in Informatics (OASIcs), Volume 70, pp. 20:1-20:16, Schloss Dagstuhl – Leibniz-Zentrum für Informatik (2019)

Copy BibTex To Clipboard

@InProceedings{sauri_et_al:OASIcs.LDK.2019.20,
  author =	{Saur{\'\i}, Roser and Mahon, Louis and Russo, Irene and Bitinis, Mironas},
  title =	{{Cross-Dictionary Linking at Sense Level with a Double-Layer Classifier}},
  booktitle =	{2nd Conference on Language, Data and Knowledge (LDK 2019)},
  pages =	{20:1--20:16},
  series =	{Open Access Series in Informatics (OASIcs)},
  ISBN =	{978-3-95977-105-4},
  ISSN =	{2190-6807},
  year =	{2019},
  volume =	{70},
  editor =	{Eskevich, Maria and de Melo, Gerard and F\"{a}th, Christian and McCrae, John P. and Buitelaar, Paul and Chiarcos, Christian and Klimek, Bettina and Dojchinovski, Milan},
  publisher =	{Schloss Dagstuhl -- Leibniz-Zentrum f{\"u}r Informatik},
  address =	{Dagstuhl, Germany},
  URL =		{https://drops.dagstuhl.de/entities/document/10.4230/OASIcs.LDK.2019.20},
  URN =		{urn:nbn:de:0030-drops-103848},
  doi =		{10.4230/OASIcs.LDK.2019.20},
  annote =	{Keywords: Word sense linking, word sense mapping, lexical translation, lexical resources, language data construction, multilingual data, data integration across languages}
}
Document
DOI: 10.4230/DagSemProc.05151.4
Argument Structure in TimeML

Authors: James Pustejovsky, Jessica Littman, and Roser Sauri

Published in: Dagstuhl Seminar Proceedings, Volume 5151, Annotating, Extracting and Reasoning about Time and Events (2005)

Abstract
TimeML is a specification language for the annotation of events and temporal expressions in natural language text. In addition, the language introduces three relational tags linking temporal objects and events to one another. These links impose both aspectual and temporal ordering over time objects, as well as mark up subordination contexts introduced by modality, evidentiality, and factivity. Given the richness of this specification, the TimeML working group decided not to include the arguments of events within the language specification itself. Full reasoning and inference over natural language texts clearly requires knowledge of events along with their participants. In this paper, we define the appropriate role of argumenthood within event markup and propose that TimeML should make a basic distinction between arguments that are events and those that are entities. We first review how TimeML treats event arguments in subordinating and aspectual contexts, creating event-event relations between predicate and argument. As it turns out, these constructions cover a large number of the argument types selected for by event predicates. We suggest that TimeML be enriched slightly to include causal predicates, such as {it lead to}, since these also involve event-event relations. We propose that all other verbal arguments be ignored by the specification, and any predicate-argument binding of participants to an event should be performed by independent means. In fact, except for the event-denoting arguments handled by the extension to TimeML proposed here, almost full temporal ordering of the events in a text can be computed without argument identification.
Cite as

James Pustejovsky, Jessica Littman, and Roser Sauri. Argument Structure in TimeML. In Annotating, Extracting and Reasoning about Time and Events. Dagstuhl Seminar Proceedings, Volume 5151, pp. 1-14, Schloss Dagstuhl – Leibniz-Zentrum für Informatik (2006)

Copy BibTex To Clipboard

@InProceedings{pustejovsky_et_al:DagSemProc.05151.4,
  author =	{Pustejovsky, James and Littman, Jessica and Sauri, Roser},
  title =	{{Argument Structure in TimeML}},
  booktitle =	{Annotating, Extracting and Reasoning about Time and Events},
  pages =	{1--14},
  series =	{Dagstuhl Seminar Proceedings (DagSemProc)},
  ISSN =	{1862-4405},
  year =	{2006},
  volume =	{5151},
  editor =	{Graham Katz and James Pustejovsky and Frank Schilder},
  publisher =	{Schloss Dagstuhl -- Leibniz-Zentrum f{\"u}r Informatik},
  address =	{Dagstuhl, Germany},
  URL =		{https://drops.dagstuhl.de/entities/document/10.4230/DagSemProc.05151.4},
  URN =		{urn:nbn:de:0030-drops-4498},
  doi =		{10.4230/DagSemProc.05151.4},
  annote =	{Keywords: Temporal annotation, event expressions, argument structure}
}

Sauri, Roser

Document
DOI: 10.4230/OASIcs.LDK.2019.20
Cross-Dictionary Linking at Sense Level with a Double-Layer Classifier

Authors: Roser Saurí, Louis Mahon, Irene Russo, and Mironas Bitinis

Published in: OASIcs, Volume 70, 2nd Conference on Language, Data and Knowledge (LDK 2019)

Abstract
We present a system for linking dictionaries at the sense level, which is part of a wider programme aiming to extend current lexical resources and to create new ones by automatic means. One of the main challenges of the sense linking task is the existence of non one-to-one mappings among senses. Our system handles this issue by addressing the task as a binary classification problem using standard Machine Learning methods, where each sense pair is classified independently from the others. In addition, it implements a second, statistically-based classification layer to also model the dependence existing among sense pairs, namely, the fact that a sense in one dictionary that is already linked to a sense in the other dictionary has a lower probability of being linked to a further sense. The resulting double-layer classifier achieves global Precision and Recall scores of 0.91 and 0.80, respectively.
Cite as

Roser Saurí, Louis Mahon, Irene Russo, and Mironas Bitinis. Cross-Dictionary Linking at Sense Level with a Double-Layer Classifier. In 2nd Conference on Language, Data and Knowledge (LDK 2019). Open Access Series in Informatics (OASIcs), Volume 70, pp. 20:1-20:16, Schloss Dagstuhl – Leibniz-Zentrum für Informatik (2019)

Copy BibTex To Clipboard

@InProceedings{sauri_et_al:OASIcs.LDK.2019.20,
  author =	{Saur{\'\i}, Roser and Mahon, Louis and Russo, Irene and Bitinis, Mironas},
  title =	{{Cross-Dictionary Linking at Sense Level with a Double-Layer Classifier}},
  booktitle =	{2nd Conference on Language, Data and Knowledge (LDK 2019)},
  pages =	{20:1--20:16},
  series =	{Open Access Series in Informatics (OASIcs)},
  ISBN =	{978-3-95977-105-4},
  ISSN =	{2190-6807},
  year =	{2019},
  volume =	{70},
  editor =	{Eskevich, Maria and de Melo, Gerard and F\"{a}th, Christian and McCrae, John P. and Buitelaar, Paul and Chiarcos, Christian and Klimek, Bettina and Dojchinovski, Milan},
  publisher =	{Schloss Dagstuhl -- Leibniz-Zentrum f{\"u}r Informatik},
  address =	{Dagstuhl, Germany},
  URL =		{https://drops.dagstuhl.de/entities/document/10.4230/OASIcs.LDK.2019.20},
  URN =		{urn:nbn:de:0030-drops-103848},
  doi =		{10.4230/OASIcs.LDK.2019.20},
  annote =	{Keywords: Word sense linking, word sense mapping, lexical translation, lexical resources, language data construction, multilingual data, data integration across languages}
}
Document
DOI: 10.4230/DagSemProc.05151.4
Argument Structure in TimeML

Authors: James Pustejovsky, Jessica Littman, and Roser Sauri

Published in: Dagstuhl Seminar Proceedings, Volume 5151, Annotating, Extracting and Reasoning about Time and Events (2005)

Abstract
TimeML is a specification language for the annotation of events and temporal expressions in natural language text. In addition, the language introduces three relational tags linking temporal objects and events to one another. These links impose both aspectual and temporal ordering over time objects, as well as mark up subordination contexts introduced by modality, evidentiality, and factivity. Given the richness of this specification, the TimeML working group decided not to include the arguments of events within the language specification itself. Full reasoning and inference over natural language texts clearly requires knowledge of events along with their participants. In this paper, we define the appropriate role of argumenthood within event markup and propose that TimeML should make a basic distinction between arguments that are events and those that are entities. We first review how TimeML treats event arguments in subordinating and aspectual contexts, creating event-event relations between predicate and argument. As it turns out, these constructions cover a large number of the argument types selected for by event predicates. We suggest that TimeML be enriched slightly to include causal predicates, such as {it lead to}, since these also involve event-event relations. We propose that all other verbal arguments be ignored by the specification, and any predicate-argument binding of participants to an event should be performed by independent means. In fact, except for the event-denoting arguments handled by the extension to TimeML proposed here, almost full temporal ordering of the events in a text can be computed without argument identification.
Cite as

James Pustejovsky, Jessica Littman, and Roser Sauri. Argument Structure in TimeML. In Annotating, Extracting and Reasoning about Time and Events. Dagstuhl Seminar Proceedings, Volume 5151, pp. 1-14, Schloss Dagstuhl – Leibniz-Zentrum für Informatik (2006)

Copy BibTex To Clipboard

@InProceedings{pustejovsky_et_al:DagSemProc.05151.4,
  author =	{Pustejovsky, James and Littman, Jessica and Sauri, Roser},
  title =	{{Argument Structure in TimeML}},
  booktitle =	{Annotating, Extracting and Reasoning about Time and Events},
  pages =	{1--14},
  series =	{Dagstuhl Seminar Proceedings (DagSemProc)},
  ISSN =	{1862-4405},
  year =	{2006},
  volume =	{5151},
  editor =	{Graham Katz and James Pustejovsky and Frank Schilder},
  publisher =	{Schloss Dagstuhl -- Leibniz-Zentrum f{\"u}r Informatik},
  address =	{Dagstuhl, Germany},
  URL =		{https://drops.dagstuhl.de/entities/document/10.4230/DagSemProc.05151.4},
  URN =		{urn:nbn:de:0030-drops-4498},
  doi =		{10.4230/DagSemProc.05151.4},
  annote =	{Keywords: Temporal annotation, event expressions, argument structure}
}
Questions / Remarks / Feedback
X

Feedback for Dagstuhl Publishing


Thanks for your feedback!

Feedback submitted

Could not send message

Please try again later or send an E-mail
© 2023-2024 Schloss Dagstuhl – LZI GmbH Imprint Privacy Contact