Abstract

Hindi-English Machine Translation is a challenging problem, owing to multiple factors including the morphological complexity and relatively free word order of Hindi, in addition to the lack of sufficient parallel training data. Neural Machine Translation (NMT) is a rapidly advancing MT paradigm and has shown promising results for many language pairs, especially in large training data scenarios. To overcome the data sparsity issue caused by the lack of large parallel corpora for Hindi-English, we propose a method to employ additional linguistic knowledge which is encoded by different phenomena depicted by Hindi. We generalize the embedding layer of the state-of-the-art Transformer model to incorporate linguistic features like POS tag, lemma and morph features to improve the translation performance. We compare the results obtained on incorporating this knowledge with the baseline systems and demonstrate significant performance improvements. Although, the Transformer NMT models have a strong efficacy to learn language constructs,we show that the usage of specific features further help in improving the translation performance.

Abstract

We propose MEE, an approach for automatic Machine Translation (MT) evaluation which leverages the similarity between embeddings of words in candidate and reference sentences to assess translation quality. Unigrams are matched based on their surface forms, root forms and meanings which aids to capture lexical, morphological and semantic equivalence. We perform experiments for MT from English to four Indian Languages (Telugu, Marathi, Bengali and Hindi) on a robust dataset comprising simple and complex sentences with good and bad translations. Further, it is observed that the proposed metric correlates better with human judgements than the existing widely used metrics.

Abstract

Discourse relations between two textual spans in a document attempt to capture the coherent structure which emerges in language use. Automatic classification of these relations remains a challenging task especially in case of implicit discourse relations, where there is no explicit textual cue which marks the discourse relation. In low resource languages, this motivates the exploration of transfer learning approaches, more particularly the cross-lingual techniques towards discourse relation classification. In this work, we explore various cross-lingual transfer techniques on Hindi Discourse Relation Bank (HDRB), a Penn Discourse Treebank styled dataset for discourse analysis in Hindi and observe performance gains in both zero shot and finetuning settings on the Hindi Discourse Relation Classification task. This is the first effort towards exploring transfer learning for Hindi Discourse relation classification to the best of our knowledge.

Abstract

This paper describes the participation of team F1toF6 (LTRC, IIIT-Hyderabad) for the WMT 2020 task, similar language translation. We experimented with attention based recurrent neural network architecture (seq2seq) for this task. We explored the use of different linguistic features like POS and Morph along with back translation for Hindi-Marathi and Marathi-Hindi machine translation.

Abstract

Word segmentation is a fundamental task for most of the NLP applications. Urdu adopts Nastalique writing style which does not have a concept of space. Furthermore, the inherent non-joining attributes of certain characters in Urdu create spaces within a word while writing in digital format. Thus, Urdu not only has space omission but also space insertion issues which make the word segmentation task challenging. In this paper, we improve upon the results of Zia, Raza and Athar (2018) by using a manually annotated corpus of 19,651 sentences along with morphological context features. Using the Conditional Random Field sequence modeler, our model achieves F1 score of 0.98 for word boundary identification and 0.92 for sub-word boundary identification tasks. The results demonstrated in this paper outperform the state-of- the-art methods.

Abstract

The rapid growth of the internet has given us a wealth of information and data spread across the web. However, as the data begins to grow we simultaneously face the grave problem of an Information Explosion. An abundance of data can lead to large scale data management problems as well as the loss of the true meaning of the data. In this paper, we present an advanced domain specific keyword extraction algorithm in order to tackle this problem of paramount importance. Our algorithm is based on a modified version of TextRank(Mihalcea and Tarau, 2004) algorithm - an algorithm based on PageRank(Page et al., 1998) to successfully determine the keywords from a domain specific document. Furthermore, this paper proposes a modification to the traditional TextRank algorithm that takes into account bigrams and trigrams and returns results with an extremely high precision. We observe how the precision and f1-score of this model outperforms other models in many domains and the recall can be easily increased by increasing the number of results without affecting the precision. We also discuss about

Abstract

The introduction of transformers in 2017 and successively BERT in 2018 brought about a revolution in the field of natural language processing. Such models are pretrained on vast amounts of data, and are easily extensible to be used for a wide variety of tasks through transfer learning. Continual work on transformer based architectures has led to a variety of new models with state of the art results. RoBERTa (Liu et al., 2019) is one such model, which brings about a series of changes to the BERT (Devlin et al., 2018) architecture and is capable of producing better quality embeddings at an expense of functionality. In this paper, we attempt to solve the well known text classification task of fine-grained domain classification using BERT and RoBERTa and perform a comparative analysis of the same. We also attempt to evaluate the impact of data preprocessing specially in the context of fine-grained domain classification. The results obtained outperformed all the other models at the ICON TechDOfication 2020 (subtask-2a) Fine-grained domain classification task and ranked first. This proves the effectiveness of our approach.

Abstract

Sentiment Analysis and other semantic tasks are commonly used for social media textual analysis to gauge public opinion and make sense from the noise on social media. The language used on social media not only commonly diverges from the formal language, but is compounded by codemixing between languages, especially in large multilingual societies like India. Traditional methods for learning semantic NLP tasks have long relied on end to end task specific training, requiring expensive data creation process, even more so for deep learning methods. This challenge is even more severe for resource scarce texts like codemixed language pairs, with lack of well learnt representations as model priors, and task specific datasets can be few and small in quantities to efficiently exploit recent deep learning approaches. To address above challenges, we introduce curriculum learning strategies for semantic tasks in code-mixed Hindi-English (Hi-En) texts, and investigate various training strategies for enhancing model performance. Our method outperforms the state of the art methods for Hi-En codemixed sentiment analysis by 3.31% accuracy, and also shows better model robustness in terms of convergence, and variance in test performance.

Abstract

This paper describes the Neural Machine Translation systems of IIIT-Hyderabad (LTRC-MT) for WAT 2019 Hindi-English shared task. We experimented with both Recurrent Neural Networks & Transformer architectures. We also show the results of our experiments of training NMT models using additional data via backtranslation.

Abstract

Automatic identification of offensive language in various social media platforms especially Twitter poses a great challenge to the AI community. The repercussions of such writings are hazardous to individuals, communities, organizations and nations. The HASOC shared task attempts for automatic detection of abusive language on Twitter in English, German and Hindi languages. As a part of this task, we (team A3-108) submitted different machine learning and neural network based models for all the languages. Our best performing model was an ensemble model of SVM, Random Forest and Adaboost classifiers with majority voting.