In this paper, we present an attention-based deep learning framework, DeepADA, which uses data augmentation to address the class imbalance problem in textual datasets. The proposed framework carries out the following functions:(i) using MPNET-based embeddings to extract keywords out of documents from the minority class, (ii) making use of a CNN-BiLSTM architecture with parallel attention to learn the important contextual words associated with the minority class documents’ keywords and provide them with word-level characteristics derived from their statistical and semantic features, (iii) using MPNET, replacing the key contextual terms derived from the oversampled documents that match to a keyword with the contextual term that best fits the context, and finally (iv) oversampling the minority class dataset to produce a balanced dataset. Using a 2-layer stacked BiLSTM classifier, we assess the efficacy of the proposed framework using the original and oversampled versions of three Amazon’s reviews datasets. We contrast the proposed data augmentation approach with two state-of-the-art text data augmentation methods. The experimental results reveal that our method produces an oversampled dataset that is more useful and helps the classifier perform better than the other two state-of-the-art methods. Nevertheless, we discover that the oversampled datasets outperformed their original ones by a wide margin.

Enabling voice assistants on small embedded devices requires a keyword spotter with a smaller model size and adequate accuracy. It becomes difficult to achieve a reasonable trade-off between a small footprint and high accuracy. Recent studies have demonstrated that convolution neural networks are also effective in the audio domain. In this paper, taking into account the nature of spectrograms, we propose a compact ResNet architecture that uses frequency-based non-square kernels to extract the maximum number of timbral features for keyword spotting. The proposed architecture is approximately three-and-a-half times smaller than a comparable architecture with conventional square kernels. On the Google’s speech command dataset v1, it outperforms both Google’s convolution neural networks and the equivalent ResNet architecture with square kernels. By implementing non-square kernels for spectrogram-related data, we can achieve a significant increase in accuracy with relatively few parameters, as compared to the conventional square kernels that are the default choice for every problem.

Self-deprecating sarcasm is a special category of sarcasm, which is nowadays popular and useful for many real-life applications, such as brand endorsement, product campaign, digital marketing, and advertisement. The self-deprecating style of campaign and marketing strategy is mainly adopted to excel brand endorsement and product sales value. In this paper, we propose an LSTM-based deep learning approach for detecting self-deprecating sarcasm in textual data. To the best of our knowledge, there is no prior work related to self-deprecating sarcasm detection using deep learning techniques. Starting with a filtering step to identify self-referential tweets, the proposed approach adopts a deep learning model using LSTM for detecting self-deprecating sarcasm. The proposed approach is evaluated over three Twitter datasets and performs significantly better in terms of precision, recall, and f-score.

Word representation using the distributional information of words from a sizeable corpus is considered efficacious in many natural language processing and text mining applications. However, distributional representation of a word is unable to capture distant relational knowledge, representing the relational semantics. In this paper, we propose a novel word representation approach using distributional and relational contexts, DRCoVe, which augments the distributional representation of a word using the relational semantics extracted as syntactic and semantic association among entities from the underlying corpus. Unlike existing approaches that use external knowledge bases representing the relational semantics for enhanced word representation, DRCoVe uses typed dependencies (aka syntactic dependencies) to extract relational knowledge from the underlying corpus. The proposed approach is applied over a biomedical text corpus to learn word representation and compared with GloVe, which is one of the most popular word embedding approaches. The evaluation results on various benchmark datasets for word similarity and word categorization tasks demonstrate the effectiveness of DRCoVe over the GloVe.