Which Kobani? A Case Study on the Role of Spatial Statistics and Semantics for Coreference Resolution Across Gazetteers

Zhu, Rui; Janowicz, Krzysztof; Yan, Bo; Hu, Yingjie

doi:10.21433/b3110080r0n8

Cited by 5 publications

(4 citation statements)

References 3 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The issue of gender bias in the Polish language concerning the detection of coreference chains was also studied (Zhu et al, 2016;Zhao et al, 2018;Kocmi et al, 2020)).…”

Section: Related Workmentioning

confidence: 99%

Proceedings of The Sixth Workshop on Computational Models of Reference, Anaphora and Coreference (CRAC 2023)

2023

View full text Add to dashboard Cite

This is the sixth edition of the Workshop on Computational Models of Reference, Anaphora and Coreference (CRAC). CRAC was first held in New Orleans five years ago in conjunction with NAACL HLT 2018. But the workshop series dates back to its predecessor, the Coreference Resolution Beyond OntoNotes (CORBON) that started in 2016, and has arguably become the primary forum for coreference researchers to present their latest results since the demise of the Discourse Anaphora and Anaphor Resolution Colloquium series in 2011. While CORBON focused on under-investigated coreference phenomena, CRAC has a broader scope, covering all cases of computational modeling of reference, anaphora, and coreference.CRAC 2023 continued to attract a large number of very high quality papers. Specifically, we received 15 submissions which were rigorously reviewed by three program committee members. Based on their recommendations, we accepted 10 papers. Two papers were withdrawn. This is the first time we are experimenting with the presentation of a non-archived work in progress. The idea is to allow authors to submit their work in progress for review. If it gets accepted, they can present the work at the workshop. However, it won't be included in the workshop proceedings. Thus, they can still submit a more complete version as original work to another venue. Overall, we were pleased with the large number of submissions as well as the quality of the accepted papers. This is the second year of the CRAC shared task on Multilingual Coreference Resolution. This allows researchers who did not participate in the workshop to disseminate their work to a smaller and more focused audience which should promote interesting discussions.We are grateful to the following people, without whom we could not have assembled an interesting program for the workshop. First, we are indebted to our program committee members. This year the reviewing load was on an average of three papers per reviewer. All of them did the incredible job of completing their reviews in a short reviewing period. This year we have two invited talks. We thank Bernd Bohnet and Milan Straka for accepting our invitation to be this year's invited speakers. We continue the tradition of having a panel on the Universal Anaphora (UA) effort-a unified, language-independent markup scheme that reflects common cross-linguistic understanding of reference-related phenomena. Motivated by Universal Dependencies, UA aims to facilitate referential analysis of the similarities and idiosyncrasies among typologically different languages, support comparative evaluation of anaphora resolution systems and enable comparative linguistic studies. Finally, we would like to thank the workshop participants for joining us in this event.We hope you will enjoy it as much as we do! -

show abstract

“…The issue of gender bias in the Polish language concerning the detection of coreference chains was also studied (Zhu et al, 2016;Zhao et al, 2018;Kocmi et al, 2020)).…”

Section: Related Workmentioning

confidence: 99%

Proceedings of The Sixth Workshop on Computational Models of Reference, Anaphora and Coreference (CRAC 2023)

2023

View full text Add to dashboard Cite

show abstract

“…The issue of gender bias in the Polish language concerning the detection of coreference chains was also studied (Zhu et al, 2016;Zhao et al, 2018;Kocmi et al, 2020)).…”

Section: Related Workmentioning

confidence: 99%

Towards Harmful Erotic Content Detection through Coreference-Driven Contextual Analysis

Okulska,

Wisnios

2023

Proceedings of the Sixth Workshop on Computational Models of Reference, Anaphora and Coreference (CRAC 2023)

View full text Add to dashboard Cite

Adult content detection still poses a great challenge for automation. Existing classifiers primarily focus on distinguishing between erotic and non-erotic texts. However, they often need more nuance in assessing the potential harm. Unfortunately, the content of this nature falls beyond the reach of generative models due to its potentially harmful nature. Ethical restrictions prohibit large language models (LLMs) from analyzing and classifying harmful erotics, let alone generating them to create synthetic datasets for other neural models. In such instances where data is scarce and challenging, a thorough analysis of the structure of such texts rather than a large model may offer a viable solution. Especially given that harmful erotic narratives, despite appearing similar to harmless ones, usually reveal their harmful nature first through contextual information hidden in the non-sexual parts of the narrative. This paper introduces a hybrid neural and rule-based context-aware system that leverages coreference resolution to identify harmful contextual cues in erotic content. Collaborating with professional moderators, we compiled a dataset and developed a classifier capable of distinguishing harmful from non-harmful erotic content. Our hybrid model, tested on Polish text, demonstrates a promising accuracy of 84% and a recall of 80%. Models based on RoBERTa and Longformer without explicit usage of coreference chains achieved significantly weaker results, underscoring the importance of coreference resolution in detecting such nuanced content as harmful erotics. This approach also offers the potential for enhanced visual explainability, supporting moderators in evaluating predictions and taking necessary actions to address harmful content.

show abstract

“…This type of coreference resolution task presents a number of challenges, most notably those discussed by Sehgal, Getoor, and Viechnicki () and Ngomo (). Existing approaches include the use of string similarity measures (Michalowski et al, ) and spatial signatures (Zhu, Janowicz, Yan, & Hu, ), to name but two. In this article, however, we focus on topological methods and the accuracy of resulting topological relations.…”

Section: Data Preparationmentioning

confidence: 99%

Computing and querying strict, approximate, and metrically refined topological relations in linked geographic data

Regalia

Janowicz

McKenzie

2019

Transactions in GIS

Self Cite

View full text Add to dashboard Cite

Geographic entities and the information associated with them play a major role in Web‐scale knowledge graphs such as Linked Data. Interestingly, almost all major datasets represent places and even entire regions as point coordinates. There are two key reasons for this. First, complex geometries are difficult to store and query using the current Linked Data technology stack to a degree where many queries take minutes to return or will simply time out. Second, the absence of complex geometries confirms a common suspicion among GIScientists, namely that for many everyday queries place‐based relational knowledge is more relevant than raw geometries alone. To give an illustrative example, the statement that the White House is in Washington, DC is more important for gaining an understating of the city than the exact geometries of both entities. This does not imply that complex geometries are unimportant but that (topological) relations should also be extracted from them. As Egenhofer and Mark (1995b) put it in their landmark paper on naive geography, topology matters, metric refines. In this work we demonstrate how to compute and utilize strict, approximate, and metrically refined topological relations between several geographic feature types in DBpedia and compare our results to approaches that compute result sets for topological queries on the fly.

show abstract

Which Kobani? A Case Study on the Role of Spatial Statistics and Semantics for Coreference Resolution Across Gazetteers

Cited by 5 publications

References 3 publications

Proceedings of The Sixth Workshop on Computational Models of Reference, Anaphora and Coreference (CRAC 2023)

Proceedings of The Sixth Workshop on Computational Models of Reference, Anaphora and Coreference (CRAC 2023)

Towards Harmful Erotic Content Detection through Coreference-Driven Contextual Analysis

Computing and querying strict, approximate, and metrically refined topological relations in linked geographic data

Contact Info

Product

Resources

About