Abstract

The aim of the current study is to propose acoustic features for detection of nasals in continuous speech. Acoustic features that represent certain characteristics of speech production are extracted. Features representing excitation source characteristcs are extracted using zero frequency filtering method. Features representing vocal tract system characteristics are extracted using zero time windowing method. Feature sets are formed by combining certain subsets of the features mentioned above. These feature sets are evaluated for their representativeness of nasals in continuous speech in three different languages, namely, English, Hindi and Telugu. Results show that nasal detection is reliable and consistent across all the languages mentioned above.

Abstract

Majority of the text modelling techniques yield only point estimates of document embeddings and lack in capturing the uncertainty of the estimates. These uncertainties give a notion of how well the embeddings represent a document. We present Bayesian subspace multinomial model (Bayesian SMM), a generative log-linear model that learns to represent documents in the form of Gaussian distributions, thereby encoding the uncertainty in its covariance. Additionally, in the proposed Bayesian SMM, we address a commonly encountered problem of intractability that appears during variational inference in mixed-logit models. We also present a generative Gaussian linear classifier for topic identification that exploits the uncertainty in document embeddings. Our intrinsic evaluation using perplexity measure shows that the proposed Bayesian SMM fits the data better as compared to variational auto-encoder based document model. Our topic identification experiments on speech (Fisher) and text (20Newsgroups) corpora show that the proposed Bayesian SMM is robust to over-fitting on unseen test data. The topic ID results show that the proposed model is significantly better than variational auto-encoder based methods and achieve similar results when compared to fully supervised discriminative models.

Abstract

In this paper we present a novel weighted NMF (WNMF) algorithm for interpolating missing data. The proposed approach has a computational cost equivalent to that of standard NMF and, additionally, has the flexibility to control the degree of interpolation in the missing data regions. Existing WNMF methods do not offer this capability and, thereby, tend to overestimate the values in the masked regions. By constraining the estimates of the missing-data regions, the proposed approach allows for a better trade-off in the interpolation. We further demonstrate the applicability of WNMF and missing data estimation to the problem of speech enhancement. In this preliminary work, we consider the improvement obtainable by applying the proposed method to ideal binary mask-based gain functions. The instrumental quality metrics (PESQ and SNR) clearly indicate the added benefit of the missing data interpolation, compared …

Abstract

Code-mixing is the phenomenon of mixing the vocabulary and syntax of multiple languages in the same sentence. It is an increasingly common occurrence in today’s multilingual society and poses a big challenge when encountered in different downstream tasks. In this paper, we present a hybrid architecture for the task of Sentiment Analysis of English-Hindi code-mixed data. Our method consists of three components, each seeking to alleviate different issues. We first generate subword level representations for the sentences using a CNN architecture. The generated representations are used as inputs to a Dual Encoder Network which consists of two different BiLSTMs - the Collective and Specific Encoder. The Collective Encoder captures the overall sentiment of the sentence, while the Specific Encoder utilizes an attention mechanism in order to focus on individual sentiment-bearing sub-words. This, combined with a Feature Network consisting of orthographic features and specially trained word embeddings, achieves state-of-the-art results - 83.54% accuracy and 0.827 F1 score - on a benchmark dataset.

Abstract

Traditional methods for deep NLG adopt pipeline approaches comprising stages such as constructing syntactic input, predicting function words, linearizing the syntactic input and generating the surface forms. Though easier to visualize, pipeline approaches suffer from error propagation. In addition, information available across modules cannot be leveraged by all modules. We construct a transition-based model to jointly perform linearization, function word prediction and morphological generation, which considerably improves upon the accuracy compared to a pipelined baseline system. On a standard deep input linearization shared task, our system achieves the best results reported so far.

Abstract

In this paper we present our work in building Natural Language Interface to Database (NLIDB) system using Intermediate query approach. This approach is demonstrated using Movie domain chatbot and can also be extended to different domains. The need of NLIDB System has increased in this fast paced world where more number of users are accessing databases through their Smart phones and web browsers. NLIDB System maps user’s Natural Language query to database query allowing user to extract information without any prior experience with databases. Results obtained are very promising and can tackle most of the user queries regarding target database.

Abstract

Social media sites like Facebook, Twitter, and other microblogging forums have emerged as a platform for people to express their opinions and views on different issues and events. It is often observed that people tend to take a stance; in favor, against or neutral towards a particular topic. The task of assessing the stance taken by the individual became significantly important with the emergence in the usage of online social platforms. Automatic stance detection system understands the user’s stance by analyzing the standalone texts against a target entity. Due to the limited contextual information a single sentence provides, it is challenging to solve this task effectively. In this paper, we introduce a Multi-Task Learning (MTL) based deep neural network architecture for automatically detecting stance present in the code-mixed corpus. We apply our approach on Hindi-English code-mixed corpus against the target entity - “Demonetisation.” Our best model achieved the result with a stance prediction accuracy of 63.2% which is a 4.5% overall accuracy improvement compared to the current supervised classification systems developed using the benchmark dataset for code-mixed data stance detection.

Abstract

We propose bilingual word embeddings based on word2vec and fastText models (CBOW and Skip-gram) to address the problem of Humor detection in Hindi-English code-mixed tweets in combination with deep learning architectures. We focus on deep learning approaches which are not widely used on code-mixed data and analyzed their performance by experimenting with three different neural network models. We propose convolution neural network (CNN) and bidirectional long-short term memory (biLSTM) (with and without Attention) models which take the generated bilingual embeddings as input. We make use of Twitter data to create bilingual word embeddings. All our proposed architectures outperform the state-of-the-art results, and Attention based bidirectional LSTM model achieved an accuracy of 73.6% which is an increment of more than 4% compared to the current stateof-the-art results.

Abstract

Anaphora resolution is a challenging task which has been the interest of NLP researchers for a long time. Traditional resolution techniques like eliminative constraints and weighted preferences were successful in many languages. However, they are ineffective in free word order languages like most South Asian languages. Heuristic and rule-based techniques were typical in these languages, which are constrained to context and domain. In this paper, we venture a new strategy using neural networks for resolving anaphora in human-human dialogues. The architecture chiefly consists of three components, a shallow parser for extracting features, a feature vector generator which produces the word embeddings, and a neural network model which will predict the antecedent mention of an anaphora. The system has been trained and tested on Telugu conversation corpus we generated. Given the advantage of the semantic information in word embeddings and appending actor, gender, number, person and part of plural features the model has reached an F1-score of 86.

Abstract

The content on social media now-a-days includes a huge number of messages both in textual and visual forms that are satirical in nature. Satire in the form of irony, sarcasm, and ridicule to a person, is depicted by memes on social media. Satire detection in text is an active area of research, but in the visual domain it is relatively less explored. The objective of our work is detection of satire in images taken from the popular photo-sharing platform - Flickr. Traditional methods for visual satire detection are based on supervised learning which has a necessary requirement of annotating the data which is a tedious task. In our work, to address this issue, we propose a novel, unsupervised approach which leverages the visual semantics of the images. We provide a study of clustering methods, where the difference between visual semantics of the two classes - satirical and non-satirical - becomes the basis for classification of visual content. Here, we suggest an autoencoder based clustering framework to effectively combine embedded feature learning and clustering assignments for detection of satire.