MétaCan
Menu
Back to cohort
Record W2611387367

Proceedings of the 2nd workshop on Multi-source, Multilingual Information Extraction and Summarization

2008· preprint· en· W2611387367 on OpenAlexaff
Thierry Poibeau, Sivaji Bandyopadhyay, Horacio Saggion, Roman Yangarber

Bibliographic record

VenueHAL (Le Centre pour la Communication Scientifique Directe) · 2008
Typepreprint
Languageen
FieldComputer Science
TopicNatural Language Processing Techniques
Canadian institutionsL'Alliance Boviteq
Fundersnot available
KeywordsAutomatic summarizationComputer scienceInformation retrievalInformation extractionExtraction (chemistry)World Wide WebChemistryChromatography
DOInot available

Abstract

fetched live from OpenAlex

Information extraction (IE) and text summarization (TS) are key technologies aiming at extracting relevant information from texts and presenting the information to the user in a condensed form. The ongoing information explosion makes IE and TS particularly critical for successful functioning within the information society. These technologies, however, face new challenges with the adoption of the Web 2.0 paradigm (e.g., blogs, wikis) due to their inherent multi-source nature. These technologies must no longer deal only with isolated texts or narratives, but with large-scale repositories or sources—possibly in several languages—containing a multiplicity of views, opinions, and commentaries on particular topics, entities and events. There is thus a need to adapt and/or develop new techniques to deal with these new phenomena. Recognising similar information across different sources and/or in different languages is of paramount importance in this multi-source, multi-lingual context. In information extraction, merging information from multiple sources can lead to increased accuracy, as compared to extraction from a single source. In text summarization, similar facts found across sources can inform sentence scoring algorithms. In question answering, the distribution of answers in similar contexts can inform answer-ranking components. Often, it is not the similarity of information that matters, but its complementary nature. In a multi-lingual context, information extraction and text summarization can provide solutions for crosslingual access: key pieces of information can be extracted from different texts in one or many languages, merged, and then conveyed in natural language in concise form. Applications need to be able to cope with the idiosyncratic nature of the new Web 2.0 media: mixed input, new jargon, ungrammatical and mixed-language input, emotional discourse, etc. In this context, synthesizing or inferring opinions from multiple sources is a new and exciting challenge for NLP. On another level, profiling of individuals who engage in the new social Web, and identifying whether a particular opinion is appropriate/relevant in a given context are important topics to be addressed. The objective of this second Multi-source Multilingual Information Extraction and Summarization (MMIES) workshop is to bring together researchers and practitioners in information-access technologies, to discuss recent approaches for dealing with multi-source and multi-lingual challenges.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.

How this classification was reachedexpand

Full frame machine prediction

Teacher imitation

Not calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.

metaresearch head score (Codex)0.010
metaresearch head score (Gemma)0.014
Version: metacan-v3-hybrid-931329e0061cValidation status: machine_predicted_unvalidated
Candidate categoriesnone
Consensus categoriesnone
DomainCandidate signal: none · Consensus signal: none
Study designCandidate signal: Not applicable · Consensus signal: none
GenreCandidate signal: Other · Consensus signal: none
Teacher disagreement score0.050
Threshold uncertainty score0.166

Distilled classifier scores by category (both heads)

CategoryCodexGemma
Metaresearch0.0100.014
Meta-epidemiology (narrow)0.0020.001
Meta-epidemiology (broad)0.0020.003
Bibliometrics0.0040.004
Science and technology studies0.0010.002
Scholarly communication0.0090.009
Open science0.0040.006
Research integrity0.0030.003
Insufficient payload (model declined to judge)0.0500.029

Machine scores (provisional)

The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.

Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.

Opus teacher head0.014
GPT teacher head0.250
Teacher spread0.236 · how far apart the two teachers sit on this one work
Validation statusscore_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from it

Classification

machine, unvalidated

Machine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.

The models applied no category: nothing in the taxonomy fit this work.
Study designNot applicable
Domainnot available
GenreOther

How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".

Quick stats

Citations0
Published2008
Admission routes1
Has abstractyes

Explore more

Same venueHAL (Le Centre pour la Communication Scientifique Directe)Same topicNatural Language Processing TechniquesFrench-language works237,207