Named Entity Recognition through Classifier Combination
By Radu Florian, Abraham Ittycheriah, Hongyan Jing, Tong Zhang. 2003
497 citations recorded by the source corpus.
This paper presents a classifier-combination experimental framework for named entity recognition in which four diverse classifiers (robust linear classifier, maximum entropy, transformation-based learning, and hidden Markov model) are combined under different conditions. When no gazetteer or other additional training resources are used, the combined system attains a performance of 91.6F on the English development data; integrating name, location and person gazetteers, and named entity systems trained on additional, more general, data reduces the F-measure error by a factor of 15 to 21% on the English data.
Subfields: Artificial Intelligence, Information Extraction
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Artificial Intelligence
- Mohit Bansal
- Luke S. Zettlemoyer
- Yoshua Bengio
- Percy Liang
- Graham Neubig
- Jianfeng Gao
- Noah A. Smith
- Christopher D. Manning
In Information Extraction
- Mark Dredze
- Jing Jiang
- Koby Crammer
- Pengfei Liu
- Jinlan Fu
- Xuanjing Huang
- Fien De Meulder
- Thien Huu Nguyen
Citations in the corpus
- Natural Language Processing (Almost) from Scratch (2011, background): “…F1 Shen and Sarkar (2005) 95.23% Sha and Pereira (2003) 94.29% Kudo and Matsumoto (2001) 93.91% (b) CHUNK System F1 Ando and Zhang (2005) 89.31% Florian et al. (2003) 88.76% Kudo and Matsumoto (2001) 88.31% (c) NER System F1 Koomen et al. (2005) 77.92% Pradhan et al. (2005) 77.30% Haghighi…”
- Introduction to the CoNLL-2003 Shared Task: Language-Independent Named Entity Recognition (2003, methodology): “Hidden Markov Models were employed by four of the systems that took part in the shared task (Florian et al., 2003; Klein et al., 2003; Mayfield et al., 2003; Whitelaw and Patrick, 2003). However, they were always used in combination with other learning techniques. Klein et al. (2003) also applied th”
- Neural Architectures for Named Entity Recognition (2016, methodology): “Next year, in the CoNLL2003 Shared Task, Florian et al. (2003) obtained the best score on German by combining the output of four diverse classifiers.”
- Multilingual Language Processing From Bytes (2015, background): “We use the 2002 and 2003 ConLL shared task datasets 6 for multilingual NER because they contain data in 4 languages (English, German, Spanish, and Dutch) with consistent annotations of named entities (PER, LOC, ORG, and MISC).”
- Bidirectional LSTM-CRF Models for Sequence Tagging (2015, background): “(Florian et al., 2003) presented the best system at the NER CoNLL 2003 challenge, with 88.76% F1 score.”
- Design Challenges and Misconceptions in Named Entity Recognition (2009, background): “It turns out that while problems of coverage and ambiguity prevent straightforward lookup, injection of gazetteer matches as features in machine-learning based approaches is critical for good performance (Cohen, 2004; Kazama and Torisawa, 2007a; Toral and Munoz, 2006; Florian et al., 2003).”
- Combining Supervised and Unsupervised Enembles for Knowledge Base Population (2016, methodology): “Ensembles have been applied to parsing (Henderson and Brill, 1999), word sense disambiguation (Pedersen, 2000), sentiment analysis (Whitehead and Yaeger, 2010) and information extraction (IE) (Florian et al., 2003; McClosky et al., 2012).”
- Empower Sequence Labeling with Task-Aware Neural Language Model (2017, methodology): “Handcrafted features were widely used in traditional methods like CRFs, HMMs, and maximum entropy classifiers (Lafferty, McCallum, and Pereira 2001; McCallum and Li 2003; Florian et al. 2003; Chieu and Ng 2002), but also make it hard to apply them to new tasks or domains.”
- A High-Performance Semi-Supervised Learning Method for Text Chunking (2005, unknown): “(Florian et al., 2003), CN03 (Chieu and Ng, 2003), KSNM03 (Klein et al., 2003).”
- Online Methods for Multi-Domain Learning and Adaptation (2008, background): “, 1997), and applications to NLP tasks (Florian et al., 2003).”
- Multi-domain learning by confidence-weighted parameter combination (2010, methodology): “As in the previous section, we were unable to obtain an update form suitable for a line search when using KL combination.”
- SpanNER: Named Entity Re-/Recognition as Span Prediction (2021, methodology): “which is achieved by (i) collecting training data by performing five-fold cross-validation (Wu et al., 2003; Florian et al., 2003) on the original training samples of each dataset (ii) training a system combiner based on collected samples.”
- Modeling Joint Entity and Relation Extraction with Table Representation (2014, background): “Local features for entities (or words) are similar to the features used by Florian et al. (2003), but some features are generalized and extended, and gazetteer features are excluded.”
- Improving Multimodal Named Entity Recognition via Entity Span Detection with Unified Multimodal Transformer (2020, background): “Traditional approaches typically focus on designing various effective NER features, followed by feeding them to different linear classifiers such as maximum entropy, conditional random fields (CRFs), and support vector machines (Chieu and Ng, 2002; Florian et al., 2003; Finkel et al., 2005; Ratinov ”
- Visual Attention Model for Name Tagging in Multimodal Social Media (2018, background): “, 2016; Ma and Hovy, 2016) proposed several neural network architectures for named tagging that outperform traditional explicit features based methods (Chieu and Ng, 2002; Florian et al., 2003; Ando and Zhang, 2005; Ratinov and Roth, 2009; Lin and Wu, 2009; Passos et al., 2014; Luo et al., 2015).”
- Semi-Supervised Structured Output Learning Based on a Hybrid Generative and Discriminative Approach (2007, unknown): “31 unlabeled data (27M words) (Ando and Zhang, 2005) (Florian et al., 2003) 88.”
- Transfer Learning for Sequence Labeling Using Source Model and Target Data (2019, methodology): “Machine learning algorithms such as Maximum Entropy, Perceptron and CRFs were typically applied (Florian et al. 2003; Chieu and Ng 2003; Diesner and Carley 2008; He and Kayaalp 2008).”