PRASHANT KAPIL - Academia.edu (original) (raw)

Papers by PRASHANT KAPIL

With the exponential rise in user-generated web content on social media, the proliferation of abu... more With the exponential rise in user-generated web content on social media, the proliferation of abusive languages towards an individual or a group across the different sections of the internet is also rapidly increasing. It is very challenging for human moderators to identify the offensive contents and filter those out. Deep neural networks have shown promise with reasonable accuracy for hate speech detection and allied applications. However, the classifiers are heavily dependent on the size and quality of the training data. Such a high-quality large data set is not easy to obtain. Moreover, the existing data sets that have emerged in recent times are not created following the same annotation guidelines and are often concerned with different types and sub-types related to hate. To solve this data sparsity problem, and to obtain more global representative features, we propose a Convolution Neural Network (CNN) based multi-task learning models (MTLs)! to leverage information from multip...

ArXiv, 2020

The phenomenal growth on the internet has helped in empowering individual's expressions, but ... more The phenomenal growth on the internet has helped in empowering individual's expressions, but the misuse of freedom of expression has also led to the increase of various cyber crimes and anti-social activities. Hate speech is one such issue that needs to be addressed very seriously as otherwise, this could pose threats to the integrity of the social fabrics. In this paper, we proposed deep learning approaches utilizing various embeddings for detecting various types of hate speeches in social media. Detecting hate speech from a large volume of text, especially tweets which contains limited contextual information also poses several practical challenges. Moreover, the varieties in user-generated data and the presence of various forms of hate speech makes it very challenging to identify the degree and intention of the message. Our experiments on three publicly available datasets of different domains shows a significant improvement in accuracy and F1-score.

This paper describes our system submitted in the shared task at COLING 2018 TRAC-1: Aggression Id... more This paper describes our system submitted in the shared task at COLING 2018 TRAC-1: Aggression Identification. The objective of this task was to predict online aggression spread through online textual post or comment. The dataset was released in two languages, English and Hindi. We submitted a single system for Hindi and a single system for English. Both the systems are based on an ensemble architecture where the individual models are based on Convoluted Neural Network and Support Vector Machine. Evaluation shows promising results for both the languages.The total submission for English was 30 and Hindi was 15. Our system on English facebook and social media obtained F1 score of 0.5151 and 0.5099 respectively where Hindi facebook and social media obtained F1 score of 0.5599 and 0.3790 respectively.

Artificial Intelligence Trends

In recent years, the increasing propagation of hate speech on social media has encouraged researc... more In recent years, the increasing propagation of hate speech on social media has encouraged researchers to address the problem of hateful content identification. To build an efficient hate speech detection model, a large number of annotated data is needed to train the model. To solve this approach we utilized eleven datasets from the hate speech domain and compared different transformer encoder-based approaches such as BERT, and ALBERT in single-task learning and multi-task learning (MTL) framework. We also leveraged the eight sentiment and emotion analysis datasets in the training to enrich the features in the MTL setting. The stacking based ensemble of BERT-MTL and ALBERT-MTL is utilized to combine the features from best two models. The experiments demonstrate the efficacy of the approach by attaining state-of-the-art results in all the datasets. The qualitative and quantitative error analysis was done to figure out the misclassified tweets and the effect of models on the different ...

Proceedings of the 13th International Workshop on Semantic Evaluation, 2019

In this paper we built several deep learning architectures to participate in shared task Of-fensE... more In this paper we built several deep learning architectures to participate in shared task Of-fensEval: Identifying and categorizing Offensive language in Social media by semEval-2019 (Zampieri et al., 2019b). The dataset was annotated with three level annotation schemes and task was to detect between offensive and not offensive, categorization and target identification in offensive contents. Deep learning models with POS information as feature were also leveraged for classification. The three best models that performed best on individual sub tasks are stacking of CNN-Bi-LSTM with Attention, BiLSTM with POS information added with word features and Bi-LSTM for third task. Our models achieved a Macro F1 score of 0.7594, 0.5378 and 0.4588 in Task(A,B,C) respectively with rank of 33 rd , 54 th and 52 nd out of 103, 75 and 65 submissions.

With the exponential rise in user-generated web content on social media, the proliferation of abu... more With the exponential rise in user-generated web content on social media, the proliferation of abusive languages towards an individual or a group across the different sections of the internet is also rapidly increasing. It is very challenging for human moderators to identify the offensive contents and filter those out. Deep neural networks have shown promise with reasonable accuracy for hate speech detection and allied applications. However, the classifiers are heavily dependent on the size and quality of the training data. Such a high-quality large data set is not easy to obtain. Moreover, the existing data sets that have emerged in recent times are not created following the same annotation guidelines and are often concerned with different types and sub-types related to hate. To solve this data sparsity problem, and to obtain more global representative features, we propose a Convolution Neural Network (CNN) based multi-task learning models (MTLs)! to leverage information from multip...

ArXiv, 2020

The phenomenal growth on the internet has helped in empowering individual's expressions, but ... more The phenomenal growth on the internet has helped in empowering individual's expressions, but the misuse of freedom of expression has also led to the increase of various cyber crimes and anti-social activities. Hate speech is one such issue that needs to be addressed very seriously as otherwise, this could pose threats to the integrity of the social fabrics. In this paper, we proposed deep learning approaches utilizing various embeddings for detecting various types of hate speeches in social media. Detecting hate speech from a large volume of text, especially tweets which contains limited contextual information also poses several practical challenges. Moreover, the varieties in user-generated data and the presence of various forms of hate speech makes it very challenging to identify the degree and intention of the message. Our experiments on three publicly available datasets of different domains shows a significant improvement in accuracy and F1-score.

This paper describes our system submitted in the shared task at COLING 2018 TRAC-1: Aggression Id... more This paper describes our system submitted in the shared task at COLING 2018 TRAC-1: Aggression Identification. The objective of this task was to predict online aggression spread through online textual post or comment. The dataset was released in two languages, English and Hindi. We submitted a single system for Hindi and a single system for English. Both the systems are based on an ensemble architecture where the individual models are based on Convoluted Neural Network and Support Vector Machine. Evaluation shows promising results for both the languages.The total submission for English was 30 and Hindi was 15. Our system on English facebook and social media obtained F1 score of 0.5151 and 0.5099 respectively where Hindi facebook and social media obtained F1 score of 0.5599 and 0.3790 respectively.

Artificial Intelligence Trends

In recent years, the increasing propagation of hate speech on social media has encouraged researc... more In recent years, the increasing propagation of hate speech on social media has encouraged researchers to address the problem of hateful content identification. To build an efficient hate speech detection model, a large number of annotated data is needed to train the model. To solve this approach we utilized eleven datasets from the hate speech domain and compared different transformer encoder-based approaches such as BERT, and ALBERT in single-task learning and multi-task learning (MTL) framework. We also leveraged the eight sentiment and emotion analysis datasets in the training to enrich the features in the MTL setting. The stacking based ensemble of BERT-MTL and ALBERT-MTL is utilized to combine the features from best two models. The experiments demonstrate the efficacy of the approach by attaining state-of-the-art results in all the datasets. The qualitative and quantitative error analysis was done to figure out the misclassified tweets and the effect of models on the different ...

Proceedings of the 13th International Workshop on Semantic Evaluation, 2019

In this paper we built several deep learning architectures to participate in shared task Of-fensE... more In this paper we built several deep learning architectures to participate in shared task Of-fensEval: Identifying and categorizing Offensive language in Social media by semEval-2019 (Zampieri et al., 2019b). The dataset was annotated with three level annotation schemes and task was to detect between offensive and not offensive, categorization and target identification in offensive contents. Deep learning models with POS information as feature were also leveraged for classification. The three best models that performed best on individual sub tasks are stacking of CNN-Bi-LSTM with Attention, BiLSTM with POS information added with word features and Bi-LSTM for third task. Our models achieved a Macro F1 score of 0.7594, 0.5378 and 0.4588 in Task(A,B,C) respectively with rank of 33 rd , 54 th and 52 nd out of 103, 75 and 65 submissions.