Topic overview

Computation and Language

Computer Vision

Explore graph Browse papers

Map preview

Topic at a glance

14115works

37927authors indexed

preprint2019arXiv

Right for the Wrong Reasons: Diagnosing Syntactic Heuristics in Natural Language Inference

A machine learning system can score well on a given test set by relying on heuristics that are effective for frequent example types but break down in more challenging cases. We study this issue within natural language inference (NLI), the task of determining whether one sentence entails another. We hypothesize that statistical NLI models may adopt three fallible syntactic heuristics: the lexical overlap heuristic, the subsequence heuristic, and the constituent heuristic. To determine whether models have adopted these heuristics, we introduce a controlled evaluation set called HANS (Heuristic Analysis for NLI Systems), which contains many examples where the heuristics fail. We find that models trained on MNLI, including BERT, a state-of-the-art model, perform very poorly on HANS, suggesting that they have indeed adopted these heuristics. We conclude that there is substantial room for improvement in NLI systems, and that the HANS dataset can motivate and measure progress in this area

preprint2018arXiv

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, 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).

preprint2017arXiv

Attention Is All You Need

The dominant sequence transduction models are based on complex recurrent or convolutional neural networks in an encoder-decoder configuration. The best performing models also connect the encoder and decoder through an attention mechanism. We propose a new simple network architecture, the Transformer, based solely on attention mechanisms, dispensing with recurrence and convolutions entirely. Experiments on two machine translation tasks show these models to be superior in quality while being more parallelizable and requiring significantly less time to train. Our model achieves 28.4 BLEU on the WMT 2014 English-to-German translation task, improving over the existing best results, including ensembles by over 2 BLEU. On the WMT 2014 English-to-French translation task, our model establishes a new single-model state-of-the-art BLEU score of 41.8 after training for 3.5 days on eight GPUs, a small fraction of the training costs of the best models from the literature. We show that the Transformer generalizes well to other tasks by applying it successfully to English constituency parsing both with large and limited training data.

preprint2017arXiv

Word Emdeddings through Hellinger PCA

Word embeddings resulting from neural language models have been shown to be successful for a large variety of NLP tasks. However, such architecture might be difficult to train and time-consuming. Instead, we propose to drastically simplify the word embeddings computation through a Hellinger PCA of the word co-occurence matrix. We compare those new word embeddings with some well-known embeddings on NER and movie review tasks and show that we can reach similar or even better performance. Although deep learning is not really necessary for generating good word embeddings, we show that it can provide an easy way to adapt embeddings to specific tasks.

preprint2017arXiv

Pyndri: a Python Interface to the Indri Search Engine

We introduce pyndri, a Python interface to the Indri search engine. Pyndri allows to access Indri indexes from Python at two levels: (1) dictionary and tokenized document collection, (2) evaluating queries on the index. We hope that with the release of pyndri, we will stimulate reproducible, open and fast-paced IR research.

preprint2016arXiv

PAMPO: using pattern matching and pos-tagging for effective Named Entities recognition in Portuguese

This paper deals with the entity extraction task (named entity recognition) of a text mining process that aims at unveiling non-trivial semantic structures, such as relationships and interaction between entities or communities. In this paper we present a simple and efficient named entity extraction algorithm. The method, named PAMPO (PAttern Matching and POs tagging based algorithm for NER), relies on flexible pattern matching, part-of-speech tagging and lexical-based rules. It was developed to process texts written in Portuguese, however it is potentially applicable to other languages as well. We compare our approach with current alternatives that support Named Entity Recognition (NER) for content written in Portuguese. These are Alchemy, Zemanta and Rembrandt. Evaluation of the efficacy of the entity extraction method on several texts written in Portuguese indicates a considerable improvement on $recall$ and $F_1$ measures.

preprint2016arXiv

Social Media Argumentation Mining: The Quest for Deliberateness in Raucousness

Argumentation mining from social media content has attracted increasing attention. The task is both challenging and rewarding. The informal nature of user-generated content makes the task dauntingly difficult. On the other hand, the insights that could be gained by a large-scale analysis of social media argumentation make it a very worthwhile task. In this position paper I discuss the motivation for social media argumentation mining, as well as the tasks and challenges involved.

preprint2017arXiv

Shortcut Sequence Tagging

Deep stacked RNNs are usually hard to train. Adding shortcut connections across different layers is a common way to ease the training of stacked networks. However, extra shortcuts make the recurrent step more complicated. To simply the stacked architecture, we propose a framework called shortcut block, which is a marriage of the gating mechanism and shortcuts, while discarding the self-connected part in LSTM cell. We present extensive empirical experiments showing that this design makes training easy and improves generalization. We propose various shortcut block topologies and compositions to explore its effectiveness. Based on this architecture, we obtain a 6% relatively improvement over the state-of-the-art on CCGbank supertagging dataset. We also get comparable results on POS tagging task.

preprint2017arXiv

Hierarchical Attention Model for Improved Machine Comprehension of Spoken Content

Multimedia or spoken content presents more attractive information than plain text content, but the former is more difficult to display on a screen and be selected by a user. As a result, accessing large collections of the former is much more difficult and time-consuming than the latter for humans. It's therefore highly attractive to develop machines which can automatically understand spoken content and summarize the key information for humans to browse over. In this endeavor, a new task of machine comprehension of spoken content was proposed recently. The initial goal was defined as the listening comprehension test of TOEFL, a challenging academic English examination for English learners whose native languages are not English. An Attention-based Multi-hop Recurrent Neural Network (AMRNN) architecture was also proposed for this task, which considered only the sequential relationship within the speech utterances. In this paper, we propose a new Hierarchical Attention Model (HAM), which constructs multi-hopped attention mechanism over tree-structured rather than sequential representations for the utterances. Improved comprehension performance robust with respect to ASR errors were

preprint2017arXiv

Self-Taught Convolutional Neural Networks for Short Text Clustering

Short text clustering is a challenging problem due to its sparseness of text representation. Here we propose a flexible Self-Taught Convolutional neural network framework for Short Text Clustering (dubbed STC^2), which can flexibly and successfully incorporate more useful semantic features and learn non-biased deep text representation in an unsupervised manner. In our framework, the original raw text features are firstly embedded into compact binary codes by using one existing unsupervised dimensionality reduction methods. Then, word embeddings are explored and fed into convolutional neural networks to learn deep feature representations, meanwhile the output units are used to fit the pre-trained binary codes in the training process. Finally, we get the optimal clusters by employing K-means to cluster the learned representations. Extensive experimental results demonstrate that the proposed framework is effective, flexible and outperform several popular clustering methods when tested on three public short text datasets.

preprint2017arXiv

Evaluating semantic models with word-sentence relatedness

Semantic textual similarity (STS) systems are designed to encode and evaluate the semantic similarity between words, phrases, sentences, and documents. One method for assessing the quality or authenticity of semantic information encoded in these systems is by comparison with human judgments. A data set for evaluating semantic models was developed consisting of 775 English word-sentence pairs, each annotated for semantic relatedness by human raters engaged in a Maximum Difference Scaling (MDS) task, as well as a faster alternative task. As a sample application of this relatedness data, behavior-based relatedness was compared to the relatedness computed via four off-the-shelf STS models: n-gram, Latent Semantic Analysis (LSA), Word2Vec, and UMBC Ebiquity. Some STS models captured much of the variance in the human judgments collected, but they were not sensitive to the implicatures and entailments that were processed and considered by the participants. All text stimuli and judgment data have been made freely available.

preprint2016arXiv

A POS Tagger for Code Mixed Indian Social Media Text - ICON-2016 NLP Tools Contest Entry from Surukam

Building Part-of-Speech (POS) taggers for code-mixed Indian languages is a particularly challenging problem in computational linguistics due to a dearth of accurately annotated training corpora. ICON, as part of its NLP tools contest has organized this challenge as a shared task for the second consecutive year to improve the state-of-the-art. This paper describes the POS tagger built at Surukam to predict the coarse-grained and fine-grained POS tags for three language pairs - Bengali-English, Telugu-English and Hindi-English, with the text spanning three popular social media platforms - Facebook, WhatsApp and Twitter. We employed Conditional Random Fields as the sequence tagging algorithm and used a library called sklearn-crfsuite - a thin wrapper around CRFsuite for training our model. Among the features we used include - character n-grams, language information and patterns for emoji, number, punctuation and web-address. Our submissions in the constrained environment,i.e., without making any use of monolingual POS taggers or the like, obtained an overall average F1-score of 76.45%, which is comparable to the 2015 winning score of 76.79%.

preprint2017arXiv

End-to-End Attention based Text-Dependent Speaker Verification

A new type of End-to-End system for text-dependent speaker verification is presented in this paper. Previously, using the phonetically discriminative/speaker discriminative DNNs as feature extractors for speaker verification has shown promising results. The extracted frame-level (DNN bottleneck, posterior or d-vector) features are equally weighted and aggregated to compute an utterance-level speaker representation (d-vector or i-vector). In this work we use speaker discriminative CNNs to extract the noise-robust frame-level features. These features are smartly combined to form an utterance-level speaker vector through an attention mechanism. The proposed attention model takes the speaker discriminative information and the phonetic information to learn the weights. The whole system, including the CNN and attention model, is joint optimized using an end-to-end criterion. The training algorithm imitates exactly the evaluation process --- directly mapping a test utterance and a few target speaker utterances into a single verification score. The algorithm can automatically select the most similar impostor for each target speaker to train the network. We demonstrated the effectiveness of

preprint2017arXiv

Labeling Topics with Images using Neural Networks

Topics generated by topic models are usually represented by lists of $t$ terms or alternatively using short phrases and images. The current state-of-the-art work on labeling topics using images selects images by re-ranking a small set of candidates for a given topic. In this paper, we present a more generic method that can estimate the degree of association between any arbitrary pair of an unseen topic and image using a deep neural network. Our method has better runtime performance $O(n)$ compared to $O(n^2)$ for the current state-of-the-art method, and is also significantly more accurate.

preprint2017arXiv

Ambiguity and Incomplete Information in Categorical Models of Language

We investigate notions of ambiguity and partial information in categorical distributional models of natural language. Probabilistic ambiguity has previously been studied using Selinger's CPM construction. This construction works well for models built upon vector spaces, as has been shown in quantum computational applications. Unfortunately, it doesn't seem to provide a satisfactory method for introducing mixing in other compact closed categories such as the category of sets and binary relations. We therefore lack a uniform strategy for extending a category to model imprecise linguistic information. In this work we adopt a different approach. We analyze different forms of ambiguous and incomplete information, both with and without quantitative probabilistic data. Each scheme then corresponds to a suitable enrichment of the category in which we model language. We view different monads as encapsulating the informational behaviour of interest, by analogy with their use in modelling side effects in computation. Previous results of Jacobs then allow us to systematically construct suitable bases for enrichment. We show that we can freely enrich arbitrary dagger compact closed cate

preprint2016arXiv

Automatic Data Deformation Analysis on Evolving Folksonomy Driven Environment

The Folksodriven framework makes it possible for data scientists to define an ontology environment where searching for buried patterns that have some kind of predictive power to build predictive models more effectively. It accomplishes this through an abstractions that isolate parameters of the predictive modeling process searching for patterns and designing the feature set, too. To reflect the evolving knowledge, this paper considers ontologies based on folksonomies according to a new concept structure called "Folksodriven" to represent folksonomies. So, the studies on the transformational regulation of the Folksodriven tags are regarded to be important for adaptive folksonomies classifications in an evolving environment used by Intelligent Systems to represent the knowledge sharing. Folksodriven tags are used to categorize salient data points so they can be fed to a machine-learning system and "featurizing" the data.

preprint2016arXiv

Intelligent information extraction based on artificial neural network

Question Answering System (QAS) is used for information retrieval and natural language processing (NLP) to reduce human effort. There are numerous QAS based on the user documents present today, but they all are limited to providing objective answers and process simple questions only. Complex questions cannot be answered by the existing QAS, as they require interpretation of the current and old data as well as the question asked by the user. The above limitations can be overcome by using deep cases and neural network. Hence we propose a modified QAS in which we create a deep artificial neural network with associative memory from text documents. The modified QAS processes the contents of the text document provided to it and find the answer to even complex questions in the documents.

preprint2017arXiv

Stance detection in online discussions

This paper describes our system created to detect stance in online discussions. The goal is to identify whether the author of a comment is in favor of the given target or against. Our approach is based on a maximum entropy classifier, which uses surface-level, sentiment and domain-specific features. The system was originally developed to detect stance in English tweets. We adapted it to process Czech news commentaries.

preprint2016arXiv

Neural Machine Translation Advised by Statistical Machine Translation

Neural Machine Translation (NMT) is a new approach to machine translation that has made great progress in recent years. However, recent studies show that NMT generally produces fluent but inadequate translations (Tu et al. 2016b; Tu et al. 2016a; He et al. 2016; Tu et al. 2017). This is in contrast to conventional Statistical Machine Translation (SMT), which usually yields adequate but non-fluent translations. It is natural, therefore, to leverage the advantages of both models for better translations, and in this work we propose to incorporate SMT model into NMT framework. More specifically, at each decoding step, SMT offers additional recommendations of generated words based on the decoding information from NMT (e.g., the generated partial translation and attention history). Then we employ an auxiliary classifier to score the SMT recommendations and a gating function to combine the SMT recommendations with NMT generations, both of which are jointly trained within the NMT architecture in an end-to-end manner. Experimental results on Chinese-English translation show that the proposed approach achieves significant and consistent improvements over state-of-the-art NMT and SMT systems

preprint2017arXiv

On (Commercial) Benefits of Automatic Text Summarization Systems in the News Domain: A Case of Media Monitoring and Media Response Analysis

In this work, we present the results of a systematic study to investigate the (commercial) benefits of automatic text summarization systems in a real world scenario. More specifically, we define a use case in the context of media monitoring and media response analysis and claim that even using a simple query-based extractive approach can dramatically save the processing time of the employees without significantly reducing the quality of their work.

preprint2017arXiv

How NOT To Evaluate Your Dialogue System: An Empirical Study of Unsupervised Evaluation Metrics for Dialogue Response Generation

We investigate evaluation metrics for dialogue response generation systems where supervised labels, such as task completion, are not available. Recent works in response generation have adopted metrics from machine translation to compare a model's generated response to a single target response. We show that these metrics correlate very weakly with human judgements in the non-technical Twitter domain, and not at all in the technical Ubuntu domain. We provide quantitative and qualitative results highlighting specific weaknesses in existing metrics, and provide recommendations for future development of better automatic evaluation metrics for dialogue systems.

preprint2016arXiv

Deep Recurrent Convolutional Neural Network: Improving Performance For Speech Recognition

A deep learning approach has been widely applied in sequence modeling problems. In terms of automatic speech recognition (ASR), its performance has significantly been improved by increasing large speech corpus and deeper neural network. Especially, recurrent neural network and deep convolutional neural network have been applied in ASR successfully. Given the arising problem of training speed, we build a novel deep recurrent convolutional network for acoustic modeling and then apply deep residual learning to it. Our experiments show that it has not only faster convergence speed but better recognition accuracy over traditional deep convolutional recurrent network. In the experiments, we compare the convergence speed of our novel deep recurrent convolutional networks and traditional deep convolutional recurrent networks. With faster convergence speed, our novel deep recurrent convolutional networks can reach the comparable performance. We further show that applying deep residual learning can boost the convergence speed of our novel deep recurret convolutional networks. Finally, we evaluate all our experimental networks by phoneme error rate (PER) with our proposed bidirectional statis

preprint2016arXiv

Write a Classifier: Predicting Visual Classifiers from Unstructured Text

People typically learn through exposure to visual concepts associated with linguistic descriptions. For instance, teaching visual object categories to children is often accompanied by descriptions in text or speech. In a machine learning context, these observations motivates us to ask whether this learning process could be computationally modeled to learn visual classifiers. More specifically, the main question of this work is how to utilize purely textual description of visual classes with no training images, to learn explicit visual classifiers for them. We propose and investigate two baseline formulations, based on regression and domain transfer, that predict a linear classifier. Then, we propose a new constrained optimization formulation that combines a regression function and a knowledge transfer function with additional constraints to predict the parameters of a linear classifier. We also propose a generic kernelized models where a kernel classifier is predicted in the form defined by the representer theorem. The kernelized models allow defining and utilizing any two RKHS (Reproducing Kernel Hilbert Space) kernel functions in the visual space and text space, respectively. We

preprint2016arXiv

The ontogeny of discourse structure mimics the development of literature

Discourse varies with age, education, psychiatric state and historical epoch, but the ontogenetic and cultural dynamics of discourse structure remain to be quantitatively characterized. To this end we investigated word graphs obtained from verbal reports of 200 subjects ages 2-58, and 676 literary texts spanning ~5,000 years. In healthy subjects, lexical diversity, graph size, and long-range recurrence departed from initial near-random levels through a monotonic asymptotic increase across ages, while short-range recurrence showed a corresponding decrease. These changes were explained by education and suggest a hierarchical development of discourse structure: short-range recurrence and lexical diversity stabilize after elementary school, but graph size and long-range recurrence only stabilize after high school. This gradual maturation was blurred in psychotic subjects, who maintained in adulthood a near-random structure. In literature, monotonic asymptotic changes over time were remarkable: While lexical diversity, long-range recurrence and graph size increased away from near-randomness, short-range recurrence declined, from above to below random levels. Bronze Age texts are structu

Inspect nearby papers, researchers, institutions and communities without opening a separate graph page.