no code implementations • NAACL 2022 • Besnik Fetahu, Anjie Fang, Oleg Rokhlenko, Shervin Malmasi
Named entity recognition (NER) in a real-world setting remains challenging and is impacted by factors like text genre, corpus quality, and data availability.
Cross-Domain Named Entity Recognition Cross-Lingual Transfer +4
no code implementations • SemEval (NAACL) 2022 • Shervin Malmasi, Anjie Fang, Besnik Fetahu, Sudipta Kar, Oleg Rokhlenko
Divided into 13 tracks, the task focused on methods to identify complex named entities (like names of movies, products and groups) in 11 languages in both monolingual and multi-lingual scenarios.
no code implementations • 25 Oct 2023 • Besnik Fetahu, Pedro Faustini, Giuseppe Castellucci, Anjie Fang, Oleg Rokhlenko, Shervin Malmasi
Using a new dataset of 6681 input questions and human written hints, we evaluated the models with automatic metrics and human evaluation.
no code implementations • 18 Sep 2023 • Hanbo Sun, Jian Gao, Xiaomin Wu, Anjie Fang, Cheng Cao, Zheng Du
Therefore, we propose HTEC for Human Transcription Error Correction.
Automatic Speech Recognition Automatic Speech Recognition (ASR) +2
no code implementations • 27 Oct 2022 • Zhiyu Chen, Jie Zhao, Anjie Fang, Besnik Fetahu, Oleg Rokhlenko, Shervin Malmasi
Furthermore, human evaluation shows that our method can generate more accurate and detailed rewrites when compared to human annotations.
no code implementations • COLING 2022 • Shervin Malmasi, Anjie Fang, Besnik Fetahu, Sudipta Kar, Oleg Rokhlenko
We present MultiCoNER, a large multilingual dataset for Named Entity Recognition that covers 3 domains (Wiki sentences, questions, and search queries) across 11 languages, as well as multilingual and code-mixing subsets.
no code implementations • NAACL 2021 • Tao Meng, Anjie Fang, Oleg Rokhlenko, Shervin Malmasi
We propose GEMNET, a novel approach for gazetteer knowledge integration, including (1) a flexible Contextual Gazetteer Representation (CGR) encoder that can be fused with any word-level model; and (2) a Mixture-of- Experts gating network that overcomes the feature overuse issue by learning to conditionally combine the context and gazetteer features, instead of assigning them fixed weights.