--- title: "Massimo Poesio — papers, page 2 — Elamur Research" description: "Massimo Poesio, Queen Mary University of London: 66 papers on Elamur Research (1997–2026), h-index 46 — affiliations, topics…" canonical: "https://research.elamur.ai/free/en/authors/massimo-poesio/page/2/" --- [Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/authors/massimo-poesio/page/2/) [home](https://md.research.elamur.ai/free/en/) / [authors](https://md.research.elamur.ai/free/en/authors/) / [Massimo Poesio](https://md.research.elamur.ai/free/en/authors/massimo-poesio/) / page 2 # Massimo Poesio · page 2 66 papers in our library · 1997–2026 · page 2 of 3 [The profile, affiliations and co-authors →](https://md.research.elamur.ai/free/en/authors/massimo-poesio/) ## Papers The most cited first (OpenAlex and our library). [← prev](https://md.research.elamur.ai/free/en/authors/massimo-poesio/) [1](https://md.research.elamur.ai/free/en/authors/massimo-poesio/) **2** [3](https://md.research.elamur.ai/free/en/authors/massimo-poesio/page/3/) [next →](https://md.research.elamur.ai/free/en/authors/massimo-poesio/page/3/) 26. **[The Effectiveness of LLMs as Annotators: A Comparative Overview and Empirical Analysis of Direct Representation](https://md.research.elamur.ai/free/en/papers/pavlovic-2024-effectiveness-llms-annotators-comparative-overview/)** *(2024 · 49 citations)* This paper asks how well large language models can replace or support human data labellers, especially for subjective tasks where annotators legitimately disagree. 27. **[The CODI-CRAC 2021 Shared Task on Anaphora, Bridging, and Discourse Deixis in Dialogue](https://md.research.elamur.ai/free/en/papers/khosla-2021-codi-crac-2021-shared-task/)** *(2021 · 47 citations)* The shared task asks systems to find anaphoric relations in conversation, a genre that OntoNotes-style corpora barely cover. 28. **[Polysemy–-Evidence from Linguistics, Behavioral Science, and Contextualized Language Models](https://md.research.elamur.ai/free/en/papers/haber-2024-polysemy-evidence-linguistics-behavioral-science/)** *(2024 · 44 citations)* This survey asks how polysemy (a word carrying several distinct but related senses, such as school meaning a building, an administration, a sports team or an institution) is mentally represented and processed, and why simple universal accounts keep falling short. 29. **[Coreference Systems Based on Kernels Methods](https://md.research.elamur.ai/free/en/papers/versley-2008-coreference-systems-based-kernels-methods/)** *(2008 · 43 citations)* The paper investigates how kernel methods can supply coreference resolution with knowledge that is partly structural and partly lexical. 30. **[ArMIS - The Arabic Misogyny and Sexism Corpus with Annotator Subjective Disagreements](https://md.research.elamur.ai/free/en/papers/almanea-2022-armis-arabic-misogyny-sexism-corpus/)** *(2022 · 42 citations)* Online misogynistic and sexist language is increasing in the Arabic-speaking world. In Saudi Arabia it has grown alongside the Vision 2030 reforms that loosened restrictions on women. Few Arabic benchmarks exist for detecting it, and those that do collapse annotator judgments into a single gold label, even though… 31. **[Merging PropBank, NomBank, TimeBank, Penn Discourse Treebank and Coreference](https://md.research.elamur.ai/free/en/papers/pustejovsky-2005-merging-propbank-nombank-timebank-penn/)** *(2005 · 38 citations)* Separate annotation projects each cover one part of semantics: PropBank and NomBank (predicate-argument structure of verbs and nouns), TimeBank (events, times and their links), the Penn Discourse Treebank (connectives as predicates), and Essex coreference (MATE scheme). 32. **[Multitask Learning-Based Neural Bridging Reference Resolution](https://md.research.elamur.ai/free/en/papers/yu-2020-multitask-learning-based-neural-bridging/)** *(2020 · 32 citations)* Bridging reference resolution connects an anaphor to an antecedent that is associated with it but not identical to it, such as "the customers" referring back to a previously mentioned supermarket. 33. **[Phrase Detectives Corpus 1.0 Crowdsourced Anaphoric Coreference.](https://md.research.elamur.ai/free/en/papers/chamberlain-2016-phrase-detectives-corpus-1-0/)** *(2016 · 31 citations)* Building coreference resolvers requires large, richly annotated data. Expert annotation is slow and expensive, so the authors explore whether a crowd of non-experts can do the job. Their data come from Phrase Detectives, an online game-with-a-purpose (GWAP). Players either propose an interpretation for a highlighted… 34. **[A Probabilistic Annotation Model for Crowdsourcing Coreference](https://md.research.elamur.ai/free/en/papers/paun-2018-probabilistic-annotation-model-crowdsourcing-coreference/)** *(2018 · 25 citations)* Large coreference corpora are costly to build with expert annotators. Crowdsourcing is a cheaper option, but existing aggregation models assume a fixed label set shared across all items. In anaphoric annotation, the possible labels (antecedents and chains) differ in every document, so those models do not apply. The… 35. **[Assessing the Capabilities of Large Language Models in Coreference: An Evaluation](https://md.research.elamur.ai/free/en/papers/gan-2024-assessing-capabilities-large-language-models/)** *(2024 · 24 citations)* The authors ask how well modern LLMs handle coreference and what this means for coreference research. They tested the GPT series (GPT3.5, GPT4, version 0613, temperature 0), the LLAMA2 family and Bard on two settings. The first was whole-document anaphora resolution on documents from the CRAC 2018 and CODI-CRAC… 36. **[Creating a Coreference Resolution System for Italian](https://md.research.elamur.ai/free/en/papers/poesio-2010-creating-coreference-resolution-system-italian/)** *(2010 · 23 citations)* The authors build what they describe as the first full-scale coreference resolution (CR) system for Italian. 37. **[Automatic Creation of Arabic Named Entity Annotated Corpus Using Wikipedia](https://md.research.elamur.ai/free/en/papers/althobaiti-2014-automatic-creation-arabic-named-entity/)** *(2014 · 22 citations)* Arabic lacks large hand-labelled corpora for named entity recognition. Its rich morphology and the absence of capital letters also make NER harder. The authors build a NE-tagged corpus, WDC, from the 26 March 2013 Arabic Wikipedia dump with almost no manual effort. 38. **[Cross-lingual Zero Pronoun Resolution](https://md.research.elamur.ai/free/en/papers/aloraini-2020-cross-lingual-zero-pronoun-resolution/)** *(2020 · 20 citations)* In pro-drop languages such as Arabic and Chinese, arguments like subjects are often left unexpressed where English would use a pronoun. 39. **[Using Automatically Extracted Minimum Spans to Disentangle Coreference Evaluation from Boundary Detection](https://md.research.elamur.ai/free/en/papers/moosavi-2019-automatically-extracted-minimum-spans-disentangle/)** *(2019 · 20 citations)* Scoring coreference by maximum spans mixes coreference errors with boundary errors such as PP attachment: «an extensive presence» and «an extensive presence, of course in this country» count as different mentions, and the system is penalised as if it had linked the wrong thing. 40. **[Towards a Corpus of Violence Acts in Arabic Social Media](https://md.research.elamur.ai/free/en/papers/alhelbawy-2016-corpus-violence-acts-arabic-social/)** *(2016 · 19 citations)* The authors build a resource to help human rights organisations find reports of abuse in Arab social media. 41. **[A Semi-supervised Learning Approach to Arabic Named Entity Recognition](https://md.research.elamur.ai/free/en/papers/althobaiti-2013-semi-supervised-learning-approach-arabic/)** *(2013 · 19 citations)* Arabic NER has mostly relied on hand-written rules or on supervised learners. Both need annotated corpora or gazetteers, which are costly to build and hard to extend to new entity categories. The authors present ASemiNER, a bootstrapping system that starts from a handful of seed names per class plus trigger nouns and… 42. **[Multi-metric optimization for coreference: The UniTN / IITP / Essex submission to the 2011 CONLL Shared Task](https://md.research.elamur.ai/free/en/papers/uryupina-2011-multi-metric-optimization-coreference-unitn/)** *(2011 · 18 citations)* Coreference resolution has no single agreed-upon evaluation metric, so the CoNLL 2011 Shared Task ranked systems by a combination of measures. 43. **[Coreference Resolution for the Biomedical Domain: A Survey](https://md.research.elamur.ai/free/en/papers/lu-2021-coreference-resolution-biomedical-domain-survey/)** *(2021 · 17 citations)* This survey looks at coreference resolution in biomedical literature. The task is a well-known bottleneck for information extraction because biomedical texts use different kinds of markables and depend on background knowledge, such as the fact that genes encode proteins, which general-purpose systems lack. The… 44. **[Stay Together: A System for Single and Split-antecedent Anaphora Resolution](https://md.research.elamur.ai/free/en/papers/yu-2021-stay-together-system-single-split/)** *(2021 · 17 citations)* Split-antecedent anaphora occurs when a plural expression refers to a set built from separately mentioned entities, as in "they" referring to John and Mary. 45. **[A Cluster Ranking Model for Full Anaphora Resolution](https://md.research.elamur.ai/free/en/papers/yu-2020-cluster-ranking-model-full-anaphora/)** *(2020 · 17 citations)* Most coreference systems are built for the CONLL 2012 (ONTONOTES) benchmark. That corpus does not annotate singletons or several kinds of non-referring expressions, such as expletives, so these systems skip part of the full anaphora resolution task. Yu, Uma and Poesio train on the ARRAU-based CRAC 2018 Shared Task… 46. **[Free the Plural: Unrestricted Split-Antecedent Anaphora Resolution](https://md.research.elamur.ai/free/en/papers/yu-2020-free-plural-unrestricted-split-antecedent/)** *(2020 · 17 citations)* Most coreference systems give a mention one antecedent, so «John met Mary. They had a chat» cannot be resolved: they refers to two separately introduced entities. The paper builds the first model for this split-antecedent case in general, on ARRAU (news, dialogue, fiction, medical; 507 training anaphors), with gold… 47. **[Domain-specific vs. Uniform Modeling for Coreference Resolution](https://md.research.elamur.ai/free/en/papers/uryupina-2012-domain-specific-vs-uniform-modeling/)** *(2012 · 17 citations)* Coreference corpora such as ACE-02, ARRAU and OntoNotes combine several domains, but each domain is fairly small. 48. **[Identifying Broken Plurals in Unvowelised Arabic Text](https://md.research.elamur.ai/free/en/papers/goweder-2004-identifying-broken-plurals-unvowelised-arabic/)** *(2004 · 17 citations)* Broken plurals (formed by changing the consonant pattern of the singular, not by adding a suffix) cannot be undone by affix stripping, so a light stemmer leaves them as separate stems. 49. **[The Universal Anaphora Scorer](https://md.research.elamur.ai/free/en/papers/yu-2022-universal-anaphora-scorer/)** *(2022 · 15 citations)* Coreference models now perform well on the identity anaphora annotated in ONTONOTES, so research is moving toward harder phenomena such as bridging, discourse deixis and plurals with split antecedents. 50. **[Extending BART to Provide a Coreference Resolution System for German](https://md.research.elamur.ai/free/en/papers/broscheit-2010-extending-bart-provide-coreference-resolution/)** *(2010 · 15 citations)* Supervised coreference resolution has mostly been developed for English. Several strong German systems exist, but none was publicly released, which makes it hard for others to start research on German. The authors address this by adapting the modular BART toolkit to German. They convert version 4 of the TüBa-D/Z… [← prev](https://md.research.elamur.ai/free/en/authors/massimo-poesio/) [1](https://md.research.elamur.ai/free/en/authors/massimo-poesio/) **2** [3](https://md.research.elamur.ai/free/en/authors/massimo-poesio/page/3/) [next →](https://md.research.elamur.ai/free/en/authors/massimo-poesio/page/3/)