DOCSLIB.ORG

Deconstructing Word Embedding Algorithms

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text
Deconstructing Word Embedding Algorithms Deconstructing word embedding algorithms Kian Kenyon-Deany∗ Edward Newell∗, Jackie Chi Kit Cheung BMO AI Capabilities Team Mila - Quebec´ AI Institute Bank of Montreal - Toronto, Ontario McGill University - Montreal,´ Quebec´ [email protected] [email protected] [email protected] Abstract find that, despite their many different hyperparam- eters, the algorithms collectively intersect upon the Word embeddings are reliable feature repre- following two key design features. First, vector- sentations of words used to obtain high qual- ity results for various NLP applications. Un- covector dot products are learned to approximate contextualized word embeddings are used in PMI statistics in the corpus. Second, modulation of many NLP tasks today, especially in resource- the loss gradient, directly or indirectly, is necessary limited settings where high memory capacity to balance weak and strong signals arising from the and GPUs are not available. Given the histor- highly imbalanced distribution of corpus statistics. ical success of word embeddings in NLP, we These findings can provide an informed basis propose a retrospective on some of the most for future development of both new embedding well-known word embedding algorithms. In algorithms and deep contextualized models. this work, we deconstruct Word2vec, GloVe, and others, into a common form, unveiling 2 Fundamental concepts some of the common conditions that seem to be required for making performant word em- We begin by formally defining embeddings, their beddings. We believe that the theoretical find- vectors and covectors (also known as “input” and ings in this paper can provide a basis for more “output” vectors (Rong, 2014; Nalisnick et al., informed development of future models. 2016)), and pointwise mutual information (PMI). 1 Introduction Embedding. In general topology, an embedding The advent of efficient uncontextualized word em- is understood as an injective structure preserving bedding algorithms (e.g., Word2vec (Mikolov et al., map, f : X ! Y , between two mathematical struc- 2013) and GloVe (Pennington et al., 2014)) marked tures X and Y . A word embedding algorithm (f) a historical breakthrough in NLP. Countless re- learns an inner-product space (Y ) to preserve a lin- searchers employed word embeddings in new mod- guistic structure within a reference corpus of text, els to improve results on a multitude of NLP prob- D (X), based on a vocabulary, V. The structure in lems. In this work, we provide a retrospective anal- D is analyzed in terms of the relationships between ysis of these groundbreaking models of the past, words induced by their co-appearances, according which simultaneously offers theoretical insights for to a certain definition of context. In such an analy- how future models can be developed and under- sis, each word figures dually: (1) as a focal element stood. We build on the theoretical work of Levy inducing a local context; and (2) as elements of the and Goldberg(2014), proving that their findings local contexts induced by focal elements. To make on the relationship between pointwise mutual in- these dual roles explicit, we distinguish two copies formation (PMI) and word embeddings go beyond of the vocabulary: the focal, or term, words VT , Word2vec and singular value decomposition. and the context words VC . In particular, we generalize several word embed- Word embedding consists of two maps: ding algorithms into a common form by proposing 1×d d×1 VC −! R VT −! R the low rank embedder framework. We decon- i 7−! hij j 7−! jji: struct each algorithm into its constituent parts, and ∗ Kian and Edward contributed equally. y This work was We use Dirac notation to distinguish vectors jji, pursued while Kian was a member of Mila. associated to focal words, from covectors hij, asso- 8479 Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing, pages 8479–8484, November 16–20, 2020. c 2020 Association for Computational Linguistics ciated to context words. In matrix notation, jji cor- fij satisfy the property: responds to a column vector and hij to a row vector. @fij Their inner product is hijji. We later demonstrate = 0 at ij = φij: (3) that many word embedding algorithms, intention- @ ij ally or not, learn a vector space where the inner The design variable φij is some function of product between a focal word j and context word corpus statistics, and its purpose is to quantita- i aims to approximate their PMI in the reference tively measure some relationship between words corpus: hijji ≈ PMI(i; j). i and j. The design variable ij is a function of model parameters that aims to approximate φ ; Pointwise mutual information (PMI). PMI is ij i.e., an embedder’s fundamental objective is to a commonly used measure of association in com- learn ≈ φ , and thus to train embeddings that putational linguistics, and has been shown to be ij ij capture the statistical relationships measured by φ . consistent and reliable for many tasks (Terra and ij The simplest choice for the kernel function , is Clarke, 2003). It measures the deviation of the ij to take = hijji. But the framework allows cooccurrence probability between two words i and ij any function that is symmetric and positive defi- j from the product of their marginal probabilities: nite, allowing the inclusion of bias parameters (e.g. pij NNij in GloVe) and subword parameterization (e.g. in PMI(i; j) := ln = ln ; (1) FastText). We later demostrate that skip-gram with pipj NiNj negative sampling takes φij := PMI(i; j) − ln k where pij is the probability of word i and word j and ij := hijji, and then learns parameter values cooccurring (for some notion of cooccurrence), and that approximate hijji ≈ PMI(i; j) − ln k. where pi and pj are marginal probabilities of words To understand the range of models encompassed, i and j occurring. The empirical PMI can be found it is helpful to see how the framework relates (but by replacing probabilities with corpus statistics. is not limited) to matrix factorization. Consider φij Words are typically considered to cooccur if they as providing the entries of a matrix: M := [φij]ij. are separated by no more than w words; Nij is For models that take ij = hijji, we can write the number of counted cooccurrences between a M^ = WV, where W is defined as having row i context i and a term j; Ni, Nj, and N are computed equal to hij, and V as having column j equal to jji. by marginalizing over the Nij statistics. Then, the loss function can be rewritten as: X 3 Word embedding algorithms L = fij (WV)ij; Mij : (i;j)2VC×VT We will now introduce the low rank embedder framework for deconstructing word embedding al- This loss function can be interpreted as matrix re- gorithms, inspired by the theory of generalized low construction error, because the constraint in Eq.3 rank models (Udell et al., 2016). We unify several means that the gradient goes to zero as WV ≈ M. word embedding algorithms by observing them all Selecting a particular low rank embedder in- from the common vantage point of their global loss stance requires key design choices to be made: we function. Note that this framework is used for theo- must chose the embedding dimension d, the form retical analysis, not necessarily implementation. of the loss terms fij, the kernel function ij, and The global loss function for a low rank embedder the association function φij. The derivative of fij takes the following form: with respect to ij, which we call the characteristic gradient, helps compare models because it exhibits X the action of the gradient yet is symmetric in the L = fij (hij; jji); φ(i; j) ; (2) parameters. In the Appendix we show how this (i;j)2VC×VT derivative relates to gradient descent. where (hij; jji) is a kernel function of learned In the following subsections, we present the @f φ(i; j) derivations of ij , , and φ for SVD (Levy model parameters, and is some scalar func- @ ij ij ij tion (such as a measure of association based on and Goldberg, 2014; Levy et al., 2015), SGNS how often i and j appear in the corpus); we denote (Mikolov et al., 2013), FastText (Joulin et al., 2017), these with ij and φij for brevity. As well, fij are GloVe (Pennington et al., 2014), and LDS (Arora loss functions that take ij and φij as inputs; all et al., 2016). The derivation for Swivel (Shazeer 8480 Model @fij φ hijji ≈ @ ij ij ij SVD 2 · ij − φij hijji PMI(i; j) PMI(i; j) − Nij SGNS (Nij + Nij ) · σ( ij) − σ(φij) hijji ln − PMI(i; j) − ln k Nij GloVe 2h(Nij) · ij − φij hijji + bi + bj ln Nij PMI(i; j) h i 2 LDS 4h(Nij) · ij − φij + C khij + jji|k ln Nij dPMI(i; j) − dγ p h i Nij · ij − φij PMI(i; j) PMI(i; j) Swivel hijji ∗ ∗ 1 · σ ij − φij PMI (i; j) PMI (i; j) Table 1: Comparison of low rank embedders. Final column shows the value of hijji at @fij = 0. GloVe and @ ij LDS set fij = 0 when Nij = 0; h(Nij) is a weighting function sublinear in Nij. Swivel takes one form when − Nij > 0 (first row) and another when Nij = 0 (second row). Nij is the number of negative samples; in SGNS, − − Nij / NiNj, and both Nij and Nij are tempered by undersampling and unigram smoothing. | et al., 2016) as a low rank embedder is trivial, as it where ij = ui Σvj. Allowing the square matrix of is already posed as a matrix factorization of PMI singular values Σ to be absorbed into the vectors (as statistics.
Load more

Recommended publications
  • Arxiv:2002.06235V1
    Semantic Relatedness and Taxonomic Word Embeddings Magdalena Kacmajor John D. Kelleher Innovation Exchange ADAPT Research Centre IBM Ireland Technological University Dublin [email protected] [email protected] Filip Klubickaˇ Alfredo Maldonado ADAPT Research Centre Underwriters Laboratories Technological University Dublin [email protected] [email protected] Abstract This paper1 connects a series of papers dealing with taxonomic word embeddings. It begins by noting that there are different types of semantic relatedness and that different lexical representa- tions encode different forms of relatedness. A particularly important distinction within semantic relatedness is that of thematic versus taxonomic relatedness. Next, we present a number of ex- periments that analyse taxonomic embeddings that have been trained on a synthetic corpus that has been generated via a random walk over a taxonomy. These experiments demonstrate how the properties of the synthetic corpus, such as the percentage of rare words, are affected by the shape of the knowledge graph the corpus is generated from. Finally, we explore the interactions between the relative sizes of natural and synthetic corpora on the performance of embeddings when taxonomic and thematic embeddings are combined. 1 Introduction Deep learning has revolutionised natural language processing over the last decade. A key enabler of deep learning for natural language processing has been the development of word embeddings. One reason for this is that deep learning intrinsically involves the use of neural network models and these models only work with numeric inputs. Consequently, applying deep learning to natural language processing first requires developing a numeric representation for words. Word embeddings provide a way of creating numeric representations that have proven to have a number of advantages over traditional numeric rep- resentations of language.
    [Show full text]
  • Combining Word Embeddings with Semantic Resources
    AutoExtend: Combining Word Embeddings with Semantic Resources Sascha Rothe∗ LMU Munich Hinrich Schutze¨ ∗ LMU Munich We present AutoExtend, a system that combines word embeddings with semantic resources by learning embeddings for non-word objects like synsets and entities and learning word embeddings that incorporate the semantic information from the resource. The method is based on encoding and decoding the word embeddings and is flexible in that it can take any word embeddings as input and does not need an additional training corpus. The obtained embeddings live in the same vector space as the input word embeddings. A sparse tensor formalization guar- antees efficiency and parallelizability. We use WordNet, GermaNet, and Freebase as semantic resources. AutoExtend achieves state-of-the-art performance on Word-in-Context Similarity and Word Sense Disambiguation tasks. 1. Introduction Unsupervised methods for learning word embeddings are widely used in natural lan- guage processing (NLP). The only data these methods need as input are very large corpora. However, in addition to corpora, there are many other resources that are undoubtedly useful in NLP, including lexical resources like WordNet and Wiktionary and knowledge bases like Wikipedia and Freebase. We will simply refer to these as resources. In this article, we present AutoExtend, a method for enriching these valuable resources with embeddings for non-word objects they describe; for example, Auto- Extend enriches WordNet with embeddings for synsets. The word embeddings and the new non-word embeddings live in the same vector space. Many NLP applications benefit if non-word objects described by resources—such as synsets in WordNet—are also available as embeddings.
    [Show full text]
  • Word-Like Character N-Gram Embedding
    Word-like character n-gram embedding Geewook Kim and Kazuki Fukui and Hidetoshi Shimodaira Department of Systems Science, Graduate School of Informatics, Kyoto University Mathematical Statistics Team, RIKEN Center for Advanced Intelligence Project fgeewook, [email protected], [email protected] Abstract Table 1: Top-10 2-grams in Sina Weibo and 4-grams in We propose a new word embedding method Japanese Twitter (Experiment 1). Words are indicated called word-like character n-gram embed- by boldface and space characters are marked by . ding, which learns distributed representations FNE WNE (Proposed) Chinese Japanese Chinese Japanese of words by embedding word-like character n- 1 ][ wwww 自己 フォロー grams. Our method is an extension of recently 2 。␣ ！！！！ 。␣ ありがと proposed segmentation-free word embedding, 3 ！␣ ありがと ][ wwww 4 .. りがとう 一个 ！！！！ which directly embeds frequent character n- 5 ]␣ ございま 微博 めっちゃ grams from a raw corpus. However, its n-gram 6 。。 うござい 什么 んだけど vocabulary tends to contain too many non- 7 ，我 とうござ 可以 うござい 8 ！！ ざいます 没有 line word n-grams. We solved this problem by in- 9 ␣我 がとうご 吗？ 2018 troducing an idea of expected word frequency. 10 了， ください 哈哈 じゃない Compared to the previously proposed meth- ods, our method can embed more words, along tion tools are used to determine word boundaries with the words that are not included in a given in the raw corpus. However, these segmenters re- basic word dictionary. Since our method does quire rich dictionaries for accurate segmentation, not rely on word segmentation with rich word which are expensive to prepare and not always dictionaries, it is especially effective when the available.
    [Show full text]
  • Learned in Speech Recognition: Contextual Acoustic Word Embeddings
    LEARNED IN SPEECH RECOGNITION: CONTEXTUAL ACOUSTIC WORD EMBEDDINGS Shruti Palaskar∗, Vikas Raunak∗ and Florian Metze Carnegie Mellon University, Pittsburgh, PA, U.S.A. fspalaska j vraunak j fmetze [email protected] ABSTRACT model [10, 11, 12] trained for direct Acoustic-to-Word (A2W) speech recognition [13]. Using this model, we jointly learn to End-to-end acoustic-to-word speech recognition models have re- automatically segment and classify input speech into individual cently gained popularity because they are easy to train, scale well to words, hence getting rid of the problem of chunking or requiring large amounts of training data, and do not require a lexicon. In addi- pre-defined word boundaries. As our A2W model is trained at the tion, word models may also be easier to integrate with downstream utterance level, we show that we can not only learn acoustic word tasks such as spoken language understanding, because inference embeddings, but also learn them in the proper context of their con- (search) is much simplified compared to phoneme, character or any taining sentence. We also evaluate our contextual acoustic word other sort of sub-word units. In this paper, we describe methods embeddings on a spoken language understanding task, demonstrat- to construct contextual acoustic word embeddings directly from a ing that they can be useful in non-transcription downstream tasks. supervised sequence-to-sequence acoustic-to-word speech recog- Our main contributions in this paper are the following: nition model using the learned attention distribution. On a suite 1. We demonstrate the usability of attention not only for aligning of 16 standard sentence evaluation tasks, our embeddings show words to acoustic frames without any forced alignment but also for competitive performance against a word2vec model trained on the constructing Contextual Acoustic Word Embeddings (CAWE).
    [Show full text]
  • Knowledge-Powered Deep Learning for Word Embedding
    Knowledge-Powered Deep Learning for Word Embedding Jiang Bian, Bin Gao, and Tie-Yan Liu Microsoft Research {jibian,bingao,tyliu}@microsoft.com Abstract. The basis of applying deep learning to solve natural language process- ing tasks is to obtain high-quality distributed representations of words, i.e., word embeddings, from large amounts of text data. However, text itself usually con- tains incomplete and ambiguous information, which makes necessity to leverage extra knowledge to understand it. Fortunately, text itself already contains well- defined morphological and syntactic knowledge; moreover, the large amount of texts on the Web enable the extraction of plenty of semantic knowledge. There- fore, it makes sense to design novel deep learning algorithms and systems in order to leverage the above knowledge to compute more effective word embed- dings. In this paper, we conduct an empirical study on the capacity of leveraging morphological, syntactic, and semantic knowledge to achieve high-quality word embeddings. Our study explores these types of knowledge to define new basis for word representation, provide additional input information, and serve as auxiliary supervision in deep learning, respectively. Experiments on an analogical reason- ing task, a word similarity task, and a word completion task have all demonstrated that knowledge-powered deep learning can enhance the effectiveness of word em- bedding. 1 Introduction With rapid development of deep learning techniques in recent years, it has drawn in- creasing attention to train complex and deep models on large amounts of data, in order to solve a wide range of text mining and natural language processing (NLP) tasks [4, 1, 8, 13, 19, 20].
    [Show full text]
  • Local Homology of Word Embeddings
    Local Homology of Word Embeddings Tadas Temcinasˇ 1 Abstract Intuitively, stratification is a decomposition of a topologi- Topological data analysis (TDA) has been widely cal space into manifold-like pieces. When thinking about used to make progress on a number of problems. stratification learning and word embeddings, it seems intu- However, it seems that TDA application in natural itive that vectors of words corresponding to the same broad language processing (NLP) is at its infancy. In topic would constitute a structure, which we might hope this paper we try to bridge the gap by arguing why to be a manifold. Hence, for example, by looking at the TDA tools are a natural choice when it comes to intersections between those manifolds or singularities on analysing word embedding data. We describe the manifolds (both of which can be recovered using local a parallelisable unsupervised learning algorithm homology-based algorithms (Nanda, 2017)) one might hope based on local homology of datapoints and show to find vectors of homonyms like ‘bank’ (which can mean some experimental results on word embedding either a river bank, or a financial institution). This, in turn, data. We see that local homology of datapoints in has potential to help solve the word sense disambiguation word embedding data contains some information (WSD) problem in NLP, which is pinning down a particular that can potentially be used to solve the word meaning of a word used in a sentence when the word has sense disambiguation problem. multiple meanings. In this work we present a clustering algorithm based on lo- cal homology, which is more relaxed1 that the stratification- 1.
    [Show full text]
  • Word Embedding, Sense Embedding and Their Application to Word Sense Induction
    Word Embeddings, Sense Embeddings and their Application to Word Sense Induction Linfeng Song The University of Rochester Computer Science Department Rochester, NY 14627 Area Paper April 2016 Abstract This paper investigates the cutting-edge techniques for word embedding, sense embedding, and our evaluation results on large-scale datasets. Word embedding refers to a kind of methods that learn a distributed dense vector for each word in a vocabulary. Traditional word embedding methods first obtain the co-occurrence matrix then perform dimension reduction with PCA. Recent methods use neural language models that directly learn word vectors by predicting the context words of the target word. Moving one step forward, sense embedding learns a distributed vector for each sense of a word. They either define a sense as a cluster of contexts where the target word appears or define a sense based on a sense inventory. To evaluate the performance of the state-of-the-art sense embedding methods, I first compare them on the dominant word similarity datasets, then compare them on my experimental settings. In addition, I show that sense embedding is applicable to the task of word sense induction (WSI). Actually we are the first to show that sense embedding methods are competitive on WSI by building sense-embedding-based systems that demonstrate highly competitive performances on the SemEval 2010 WSI shared task. Finally, I propose several possible future research directions on word embedding and sense embedding. The University of Rochester Computer Science Department supported this work. Contents 1 Introduction 3 2 Word Embedding 5 2.1 Skip-gramModel................................
    [Show full text]
  • Learning Word Meta-Embeddings by Autoencoding
    Learning Word Meta-Embeddings by Autoencoding Cong Bao Danushka Bollegala Department of Computer Science Department of Computer Science University of Liverpool University of Liverpool [email protected] [email protected] Abstract Distributed word embeddings have shown superior performances in numerous Natural Language Processing (NLP) tasks. However, their performances vary significantly across different tasks, implying that the word embeddings learnt by those methods capture complementary aspects of lexical semantics. Therefore, we believe that it is important to combine the existing word em- beddings to produce more accurate and complete meta-embeddings of words. We model the meta-embedding learning problem as an autoencoding problem, where we would like to learn a meta-embedding space that can accurately reconstruct all source embeddings simultaneously. Thereby, the meta-embedding space is enforced to capture complementary information in differ- ent source embeddings via a coherent common embedding space. We propose three flavours of autoencoded meta-embeddings motivated by different requirements that must be satisfied by a meta-embedding. Our experimental results on a series of benchmark evaluations show that the proposed autoencoded meta-embeddings outperform the existing state-of-the-art meta- embeddings in multiple tasks. 1 Introduction Representing the meanings of words is a fundamental task in Natural Language Processing (NLP). A popular approach to represent the meaning of a word is to embed it in some fixed-dimensional
    [Show full text]
  • A Comparison of Word Embeddings and N-Gram Models for Dbpedia Type and Invalid Entity Detection †
    information Article A Comparison of Word Embeddings and N-gram Models for DBpedia Type and Invalid Entity Detection † Hanqing Zhou *, Amal Zouaq and Diana Inkpen School of Electrical Engineering and Computer Science, University of Ottawa, Ottawa ON K1N 6N5, Canada; [email protected] (A.Z.); [email protected] (D.I.) * Correspondence: [email protected]; Tel.: +1-613-562-5800 † This paper is an extended version of our conference paper: Hanqing Zhou, Amal Zouaq, and Diana Inkpen. DBpedia Entity Type Detection using Entity Embeddings and N-Gram Models. In Proceedings of the International Conference on Knowledge Engineering and Semantic Web (KESW 2017), Szczecin, Poland, 8–10 November 2017, pp. 309–322. Received: 6 November 2018; Accepted: 20 December 2018; Published: 25 December 2018 Abstract: This article presents and evaluates a method for the detection of DBpedia types and entities that can be used for knowledge base completion and maintenance. This method compares entity embeddings with traditional N-gram models coupled with clustering and classification. We tackle two challenges: (a) the detection of entity types, which can be used to detect invalid DBpedia types and assign DBpedia types for type-less entities; and (b) the detection of invalid entities in the resource description of a DBpedia entity. Our results show that entity embeddings outperform n-gram models for type and entity detection and can contribute to the improvement of DBpedia’s quality, maintenance, and evolution. Keywords: semantic web; DBpedia; entity embedding; n-grams; type identification; entity identification; data mining; machine learning 1. Introduction The Semantic Web is defined by Berners-Lee et al.
    [Show full text]
  • Arxiv:2007.00183V2 [Eess.AS] 24 Nov 2020
    WHOLE-WORD SEGMENTAL SPEECH RECOGNITION WITH ACOUSTIC WORD EMBEDDINGS Bowen Shi, Shane Settle, Karen Livescu TTI-Chicago, USA fbshi,settle.shane,[email protected] ABSTRACT Segmental models are sequence prediction models in which scores of hypotheses are based on entire variable-length seg- ments of frames. We consider segmental models for whole- word (“acoustic-to-word”) speech recognition, with the feature vectors defined using vector embeddings of segments. Such models are computationally challenging as the number of paths is proportional to the vocabulary size, which can be orders of magnitude larger than when using subword units like phones. We describe an efficient approach for end-to-end whole-word segmental models, with forward-backward and Viterbi de- coding performed on a GPU and a simple segment scoring function that reduces space complexity. In addition, we inves- tigate the use of pre-training via jointly trained acoustic word embeddings (AWEs) and acoustically grounded word embed- dings (AGWEs) of written word labels. We find that word error rate can be reduced by a large margin by pre-training the acoustic segment representation with AWEs, and additional Fig. 1. Whole-word segmental model for speech recognition. (smaller) gains can be obtained by pre-training the word pre- Note: boundary frames are not shared. diction layer with AGWEs. Our final models improve over segmental models, where the sequence probability is com- prior A2W models. puted based on segment scores instead of frame probabilities. Index Terms— speech recognition, segmental model, Segmental models have a long history in speech recognition acoustic-to-word, acoustic word embeddings, pre-training research, but they have been used primarily for phonetic recog- nition or as phone-level acoustic models [11–18].
    [Show full text]
  • Arxiv:2104.05274V2 [Cs.CL] 27 Aug 2021 NLP field Due to Its Excellent Performance in Various Tasks
    Learning to Remove: Towards Isotropic Pre-trained BERT Embedding Yuxin Liang1, Rui Cao1, Jie Zheng?1, Jie Ren2, and Ling Gao1 1 Northwest University, Xi'an, China fliangyuxin,[email protected], fjzheng,[email protected] 2 Shannxi Normal University, Xi'an, China [email protected] Abstract. Research in word representation shows that isotropic em- beddings can significantly improve performance on downstream tasks. However, we measure and analyze the geometry of pre-trained BERT embedding and find that it is far from isotropic. We find that the word vectors are not centered around the origin, and the average cosine similar- ity between two random words is much higher than zero, which indicates that the word vectors are distributed in a narrow cone and deteriorate the representation capacity of word embedding. We propose a simple, and yet effective method to fix this problem: remove several dominant directions of BERT embedding with a set of learnable weights. We train the weights on word similarity tasks and show that processed embed- ding is more isotropic. Our method is evaluated on three standardized tasks: word similarity, word analogy, and semantic textual similarity. In all tasks, the word embedding processed by our method consistently out- performs the original embedding (with average improvement of 13% on word analogy and 16% on semantic textual similarity) and two baseline methods. Our method is also proven to be more robust to changes of hyperparameter. Keywords: Natural language processing · Pre-trained embedding · Word representation · Anisotropic. 1 Introduction With the rise of Transformers [12], its derivative model BERT [1] stormed the arXiv:2104.05274v2 [cs.CL] 27 Aug 2021 NLP field due to its excellent performance in various tasks.
    [Show full text]
  • A Survey of Embeddings in Clinical Natural Language Processing Kalyan KS, S
    This paper got published in Journal of Biomedical Informatics. Updated version is available @ https://doi.org/10.1016/j.jbi.2019.103323 SECNLP: A Survey of Embeddings in Clinical Natural Language Processing Kalyan KS, S. Sangeetha Text Analytics and Natural Language Processing Lab Department of Computer Applications National Institute of Technology Tiruchirappalli, INDIA. [email protected], [email protected] ABSTRACT Traditional representations like Bag of words are high dimensional, sparse and ignore the order as well as syntactic and semantic information. Distributed vector representations or embeddings map variable length text to dense fixed length vectors as well as capture prior knowledge which can transferred to downstream tasks. Even though embedding has become de facto standard for representations in deep learning based NLP tasks in both general and clinical domains, there is no survey paper which presents a detailed review of embeddings in Clinical Natural Language Processing. In this survey paper, we discuss various medical corpora and their characteristics, medical codes and present a brief overview as well as comparison of popular embeddings models. We classify clinical embeddings into nine types and discuss each embedding type in detail. We discuss various evaluation methods followed by possible solutions to various challenges in clinical embeddings. Finally, we conclude with some of the future directions which will advance the research in clinical embeddings. Keywords: embeddings, distributed representations, medical, natural language processing, survey 1. INTRODUCTION Distributed vector representation or embedding is one of the recent as well as prominent addition to modern natural language processing. Embedding has gained lot of attention and has become a part of NLP researcher’s toolkit.
    [Show full text]