Named Entity Recognition with a Maximum Entropy Approach

By Hai Leong Chieu, Hwee Tou Ng. 2003

275 citations recorded by the source corpus.

The named entity recognition (NER) task involves identifying noun phrases that are names, and assigning a class to each name. This task has its origin from the Message Understanding Conferences (MUC) in the 1990s, a series of conferences aimed at evaluating systems that extract information from natural language texts. It became evident that in order to achieve good performance in information extraction, a system needs to be able to recognize names. A separate subtask on NER was created in MUC-6 and MUC-7 (Chinchor, 1998).

Subfields: Artificial Intelligence

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

Citations in the corpus

Bidirectional LSTM-CRF Models for Sequence Tagging (2015, background): “The second best performer of CoNLL 2003 (Chieu., 2003) was 88.”
Design Challenges and Misconceptions in Named Entity Recognition (2009, methodology): “(Chieu and Ng, 2003) used features that aggregate, for each document, the context tokens appear in. Sample features are: the longest capitilized sequence of words in the document which contains the current token and the token appears before a company marker such as ltd. elsewhere in text .”
Introduction to the CoNLL-2005 Shared Task: Semantic Role Labeling (2005, background): “• Named Entities predicted with the MaximumEntropy based tagger of Chieu and Ng (2003). The tagger follows the CoNLL-2003 task setting (Tjong Kim Sang and De Meulder, 2003), and thus is not developed with WSJ data.”
A Brief Survey of Text Mining: Classification, Clustering and Extraction Techniques (2017, methodology): “Most of the named entity recognition techniques are statistical learning methods such as hidden Markov models [13], maximum entropy models [29], support vector machines [69] and”
A Framework for Learning Predictive Structures from Multiple Tasks and Unlabeled Data (2005, unknown): “7M-word labeled data CN03(Chieu and Ng, 2003) 88.”
Phrase Clustering for Discriminative Learning (2009, background): “88.44 +4.66 (Suzuki and Isozaki, 2008) 89.92 (Ando and Zhang, 2005) 89.31 (Florian et al., 2003) 88.76 (Chieu and Ng, 2003) 88.31 (Klein et al., 2003) 86.31 Given an input x , represented as a vector of m features: ( x 1 , x 2 , ....., x m ), a logistic regression classifier with parameter vector…”
An Introduction to Conditional Random Fields (2010, unknown): “For this particular task, Chieu and Ng [20] had an especially good set of features, during the competition was the best result obtained by a single model (i.”
Mining knowledge from text using information extraction (2005, methodology): “Using this general approach, IE systems have been developed that use many different trained classifiers such as decision trees [3], boosting [15], memory-based learning (MBL) [43], support-vector machines (SVMs) [40], maximum entropy (MaxEnt) [17], transformation-based learning (TBL)[68] and many ot”
Semi-Supervised Structured Output Learning Based on a Hybrid Generative and Discriminative Approach (2007, background): “…ASO-semi 89.31 unlabeled data (27M words) (Ando and Zhang, 2005) (Florian et al., 2003) 88.76 their own large gazetteers, 2M-word labeled data (Chieu and Ng, 2003) 88 (CoNLL-2000) experiments 4 indicate that HySOL is rather robust with respect to the hyper-parameter since we can obtain…”