Recent methods for learning vector space representations of words have succeeded in capturing fine-grained semantic and syntactic regularities using vector arithmetic, but the origin of these regularities has remained opaque. We analyze and make explicit the model properties needed for such regularities to emerge in word vectors. The result is a new global logbilinear regression model that combines the advantages of the two major model families in the literature: global matrix factorization and local context window methods. Our model efficiently leverages statistical information by training only on the nonzero elements in a word-word cooccurrence matrix, rather than on the entire sparse matrix or on individual context windows in a large corpus. The model produces a vector space with meaningful substructure, as evidenced by its performance of 75% on a recent word analogy task. It also outperforms related models on similarity tasks and named entity recognition.

/pdf/glove-global-vectors-for-word-representation-55l84hyy3b.pdf

Glove: Global Vectors for Word Representation

We introduce a new language representation model called BERT, which stands for Bidirectional Encoder Representations from Transformers. Unlike recent language representation models, BERT is designed to pre-train deep bidirectional representations from unlabeled text by jointly conditioning on both left and right context in all layers. As a result, the pre-trained BERT model can be fine-tuned with just one additional output layer to create state-of-the-art models for a wide range of tasks, such as question answering and language inference, without substantial task-specific architecture modifications. 
BERT is conceptually simple and empirically powerful. It obtains new state-of-the-art results on eleven natural language processing tasks, including pushing the GLUE score to 80.5% (7.7% point absolute improvement), MultiNLI accuracy to 86.7% (4.6% absolute improvement), SQuAD v1.1 question answering Test F1 to 93.2 (1.5 point absolute improvement) and SQuAD v2.0 Test F1 to 83.1 (5.1 point absolute improvement).

BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding

We introduce a new language representation model called BERT, which stands for Bidirectional Encoder Representations from Transformers. Unlike recent language representation models (Peters et al., 2018a; Radford et al., 2018), BERT is designed to pre-train deep bidirectional representations from unlabeled text by jointly conditioning on both left and right context in all layers. As a result, the pre-trained BERT model can be fine-tuned with just one additional output layer to create state-of-the-art models for a wide range of tasks, such as question answering and language inference, without substantial task-specific architecture modifications. BERT is conceptually simple and empirically powerful. It obtains new state-of-the-art results on eleven natural language processing tasks, including pushing the GLUE score to 80.5 (7.7 point absolute improvement), MultiNLI accuracy to 86.7% (4.6% absolute improvement), SQuAD v1.1 question answering Test F1 to 93.2 (1.5 point absolute improvement) and SQuAD v2.0 Test F1 to 83.1 (5.1 point absolute improvement).

We introduce a new type of deep contextualized word representation that models both (1) complex characteristics of word use (e.g., syntax and semantics), and (2) how these uses vary across linguistic contexts (i.e., to model polysemy). Our word vectors are learned functions of the internal states of a deep bidirectional language model (biLM), which is pre-trained on a large text corpus. We show that these representations can be easily added to existing models and significantly improve the state of the art across six challenging NLP problems, including question answering, textual entailment and sentiment analysis. We also present an analysis showing that exposing the deep internals of the pre-trained network is crucial, allowing downstream models to mix different types of semi-supervision signals.

/pdf/deep-contextualized-word-representations-11iek0hsjp.pdf

Deep contextualized word representations

Comunicacio presentada a la 2016 Conference of the North American Chapter of the Association for Computational Linguistics, celebrada a San Diego (CA, EUA) els dies 12 a 17 de juny 2016.

/pdf/neural-architectures-for-named-entity-recognition-4xswt97nv2.pdf

Neural Architectures for Named Entity Recognition

We describe the CoNLL-2003 shared task: language-independent named entity recognition. We give background information on the data sets (English and German) and the evaluation method, present a general overview of the systems that have taken part in the task and discuss their performance.

/pdf/introduction-to-the-conll-2003-shared-task-language-2wa2vxyakg.pdf

Introduction to the CoNLL-2003 shared task: language-independent named entity recognition

Introduction to the CoNLL-2003 Shared Task: Language-Independent Named Entity Recognition

Introduction to the CoNLL-2003 shared task

Comparative machine learning experiments have become an important methodology in empirical approaches to natural language processing (i) to investigate which machine learning algorithms have the 'right bias' to solve specific natural language processing tasks, and (ii) to investigate which sources of information add to accuracy in a learning approach. Using automatic word sense disambiguation as an example task, we show that with the methodology currently used in comparative machine learning experiments, the results may often not be reliable because of the role of and interaction between feature selection and algorithm parameter optimization. We propose genetic algorithms as a practical approach to achieve both higher accuracy within a single approach, and more reliable comparisons.

https://biblio.ugent.be/publication/598083/file/601127.pdf

Combined optimization of feature selection and algorithm parameters in machine learning of language

We used the memory-based learner Timbl (Daelemans et al., 2002) to find names in English and German newspaper text. A first system used only the training data, and a number of gazetteers. The results show that gazetteers are not beneficial in the English case, while they are for the German data. Type-token generalization was applied, but also reduced performance. The second system used gazetteers derived from the unannotated corpus, as well as the ratio of capitalized versus uncapitalized use of each word. These strategies gave an increase in performance.

/pdf/memory-based-named-entity-recognition-using-unannotated-data-3wjtuh4c7f.pdf

Fien De Meulder

Papers

Introduction to the CoNLL-2003 shared task: language-independent named entity recognition

Introduction to the CoNLL-2003 Shared Task: Language-Independent Named Entity Recognition

Introduction to the CoNLL-2003 shared task

Combined optimization of feature selection and algorithm parameters in machine learning of language

Memory-based named entity recognition using unannotated data