Problem Of Named Entity Recognition Research Articles

Arabic named entity recognition via deep co-learning

Named entity recognition (NER) is an important natural language processing (NLP) task with many applications. We tackle the problem of Arabic NER using deep learning based on Arabic word embeddings that capture syntactic and semantic relationships between words. Deep learning has been shown to perform significantly better than other approaches for various NLP tasks including NER. However, deep-learning models also require a significantly large amount of training data, which is highly lacking in the case of the Arabic language. To remedy this, we adopt the semi-supervised co-training approach to the realm of deep learning, which we refer to as deep co-learning. Our deep co-learning approach makes use of a small amount of labeled data, which is augmented with partially labeled data that is automatically generated from Wikipedia. Our approach relies only on word embeddings as features and does not involve any additional feature engineering. Nonetheless, when tested on three different Arabic NER benchmarks, our approach consistently outperforms state-of-the-art Arabic NER approaches, including ones that employ carefully-crafted NLP features. It also consistently outperforms various baselines including purely-supervised deep-learning approaches as well as semi-supervised ones that make use of only unlabeled data such as self-learning and the traditional co-training approach.

A New Efficiency Approach For Chinese Litigants Extraction

The judgment document is the carrier of the result of the litigious activity, with complete structure and complete elements. Extracting the information of the entity information of the plaintiff, the defendant, the third party and other litigants in the judgment document is of great significance to establishing the legal dispute view of the litigants, understanding its mutual relations and the economic credit status and providing decision support for enterprise or individual production and business activities. Litigants extraction can be regarded as the named entity recognition problem. The existing open domain entity recognition model is not effective for the entity recognition of such text and can’t meet the application requirements. This paper proposes a new joint named entity recognition method (Joint Method for Litigants Extraction, JMLE), which can make full use of the information of the text itself. By combining the common named entity recognition method, it realizes the efficient information extraction of the litigants in the judgment document without hand-annotated corpus. Experiments show that the method proposed in this paper can achieve a better performance for the information extraction of the litigants in the judgment document, and the F1 value can reach 89.37%.

Named Entity Recognition with Word Embeddings and Wikipedia Categories for a Low-Resource Language

In this article, we propose a word embedding--based named entity recognition (NER) approach. NER is commonly approached as a sequence labeling task with the application of methods such as conditional random field (CRF). However, for low-resource languages without the presence of sufficiently large training data, methods such as CRF do not perform well. In our work, we make use of the proximity of the vector embeddings of words to approach the NER problem. The hypothesis is that word vectors belonging to the same name category, such as a person’s name, occur in close vicinity in the abstract vector space of the embedded words. Assuming that this clustering hypothesis is true, we apply a standard classification approach on the vectors of words to learn a decision boundary between the NER classes. Our NER experiments are conducted on a morphologically rich and low-resource language, namely Bengali. Our approach significantly outperforms standard baseline CRF approaches that use cluster labels of word embeddings and gazetteers constructed from Wikipedia. Further, we propose an unsupervised approach (that uses an automatically created named entity (NE) gazetteer from Wikipedia in the absence of training data). For a low-resource language, the word vectors obtained from Wikipedia are not sufficient to train a classifier. As a result, we propose to make use of the distance measure between the vector embeddings of words to expand the set of Wikipedia training examples with additional NEs extracted from a monolingual corpus that yield significant improvement in the unsupervised NER performance. In fact, our expansion method performs better than the traditional CRF-based (supervised) approach (i.e., F-score of 65.4% vs. 64.2%). Finally, we compare our proposed approach to the official submission for the IJCNLP-2008 Bengali NER shared task and achieve an overall improvement of F-score 11.26% with respect to the best official system.

Biomedical named entity recognition using generalized expectation criteria

It is difficult to apply machine learning to a domain which is short of labeled training data, such as biomedical named entity recognition (NER) which remains a challenging task because of its extraordinary complex nomenclature. In this paper, we proposed a semi-supervised method which can train condition random field (CRF) models using generalized expectation (GE) criteria to solve biomedical named entity recognition problem. In the proposed method, instead of “instance” labeling, the “feature” labeling is applied to get the training data which can save lots of labeling time. Latent Dirichlet Allocation (LDA) model was involved to choose the features for labeling. Experiment results show that the proposed method can dramatically improve the performance of biomedical NER through incorporating unlabeled data by feature labeling.

Named entity recognition from Arabic Wikipedia

Abstract Named Entity Recognition (NER) is the problem of locating mentions to entities such as persons, locations and organizations. The named entity information is helpful for reducing the complexity of monolingual and multilingual processing tasks, such as information extraction, parsing and machine translation. We investigate the Arabic NER problem from the Arabic Wikipedia text. We employ statistical sequence labeling methods for solving the NER task. Previous studies suggest that sequence labeling methods, such as Conditional Random Fields, are the state of the art NER frameworks. The sequence labeling methods require human labeled training data. Most ofthe Arabic human labeled data for NER belong to the political news domain and the consequent trained models are biased towards the news domain. In contrast, our target test data (Arabic Wikipedia articles) has a very diverse set of topics. The domain mismatch between the train and test data results in poor NER performance. In order to reduce the coverage problem, we present three techniques: (1) we use the Wikipedia network structure to collect additional information about the text. Information such as monolingual and cross-lingual hyperlinks and text formatting lead us to use new features of the Wikipedia text in NER models. Moreover, we use cross-lingual projection to collect named entity information from English Wikipedia. (2) We use a domain adaptation technique to shift the model from the baseline political domain to domains relevant to our test data. Our model adaptation uses a small set of in-house-labeled Arabic Wikipedia articles. (3) We use self-training to port from a fully supervised to a semi-supervised learning framework: we collect a large volume of unlabeled Arabic Wikipedia articles to expand the underlying NER domain to new text domains. Our model expansion is gradual and iterative. In each iteration we add a new set of unlabeled articles to the training and use the current model to label and construct a larger model. Our NER evaluations are based on the standard precision and recall metrics.We evaluate our proposed framework in four different text domains ofArabic Wikipedia.

Scalable biomedical Named Entity Recognition: investigation of a database-supported SVM approach

This paper explores scalability issues associated with the Named Entity Recognition problem in the biomedical publications domain using Support Vector Machines. The performance results using existing binary and multi-class SVMs with increasing training data are compared to results obtained using our new implementations. Our approach eliminates prior language or domain-specific knowledge and achieves good out-of-the-box accuracy measures comparable to those obtained using more complex approaches. The training time of multi-class SVMs is reduced by several orders of magnitude, which would make support vector machines a more viable and practical solution for real-world problems with large datasets.

Machine learning-based named entity recognition via effective integration of various evidences

Named entity recognition identifies and classifies entity names in a text document into some predefined categories. It resolves the “who”, “where” and “how much” problems in information extraction and leads to the resolution of the “what” and “how” problems in further processing. This paper presents a Hidden Markov Model (HMM) and proposes a HMM-based named entity recognizer implemented as the system PowerNE. Through the HMM and an effective constraint relaxation algorithm to deal with the data sparseness problem, PowerNE is able to effectively apply and integrate various internal and external evidences of entity names. Currently, four evidences are included: (1) a simple deterministic internal feature of the words, such as capitalization and digitalization; (2) an internal semantic feature of the important triggers; (3) an internal gazetteer feature, which determines the appearance of the current word string in the provided gazetteer list; and (4) an external macro context feature, which deals with the name alias phenomena. In this way, the named entity recognition problem is resolved effectively. PowerNE has been benchmarked with the Message Understanding Conferences (MUC) data. The evaluation shows that, using the formal training and test data of the MUC-6 and MUC-7 English named entity tasks, and it achieves the F-measures of 96.6 and 94.1, respectively. Compared with the best reported machine learning system, it achieves a 1.7 higher F-measure with one quarter of the training data on MUC-6, and a 3.6 higher F-measure with one ninth of the training data on MUC-7. In addition, it performs slightly better than the best reported handcrafted rule-based systems on MUC-6 and MUC-7.