Code-switching is a common phenomenon in multilingual communities and is often used on social media. However, sentiment analysis of code-switched data is a challenging yet less explored area of research. This paper aims to develop a sentiment analysis system for code-switched data. In this paper, we present a novel approach combining two transformers using logits of their output and feeding them to a neural network for classification. We show the efficacy of our approach using two benchmark datasets, viz., English-Hindi (En-Hi), and English-Spanish (En-Es) availed by Microsoft GLUECoS. Our approach results in an F1 score of 73.66% for En-Es and 61.24% for En-Hi, significantly higher than the best model reported for the GLUECoS benchmark dataset.

Initially, we analyzed the datasets in a statistical way so as to learn about various sections’ contributions to the final summary in both the pros and life datasets. We found that both the datasets have an Introduction and Abstract along with some initial parts of the results contributing to the summary. We considered only these sections in the next stage of analysis. We found the optimal length or no of sentences of each of the Introduction, abstract, and result which contributes best to the summary. After this statistical analysis, we took the pre-trained model Facebook/bart-base and fine-tuned it with both the datasets PLOS and eLife. While fine-tuning and testing the results we have used chunking because the text lengths are huge. So to not lose information due to the number of token constraints of the model, we used chunking. Finally, we saw the eLife model giving more accurate results than PLOS in terms of readability aspect, probably because the PLOS summary is closer to its abstract, we have considered the eLife model as our final model and tuned the hyperparameters. We are ranked 7th overall and 1st in readability

This paper describes the contribution of team PHG to the WASSA 2022 shared task on Empathy Prediction and Emotion Classification. The broad goal of this task was to model an empathy score, a distress score and the type of emotion associated with the person who had reacted to the essay written in response to a newspaper article. We have used the RoBERTa model for training and top of which few layers are added to finetune the transformer. We also use few machine learning techniques to augment as well as upsample the data. Our system achieves a Pearson Correlation Coefficient of 0.488 on Task 1 (Empathy - 0.470 and Distress - 0.506) and Macro F1-score of 0.531 on Task 2.

The last few years have witnessed an exponential rise in the propagation of offensive text on social media. Identification of this text with high precision is crucial for the well-being of society. Most of the existing approaches tend to give high toxicity scores to innocuous statements (e.g., “I am a gay man”). These false positives result from over-generalization on the training data where specific terms in the statement may have been used in a pejorative sense (e.g., “gay”). Emphasis on such words alone can lead to discrimination against the classes these systems are designed to protect. In this paper, we address the problem of offensive language detection on Twitter, while also detecting the type and the target of the offense. We propose a novel approach called SyLSTM, which integrates syntactic features in the form of the dependency parse tree of a sentence and semantic features in the form of word embeddings into a deep learning architecture using a Graph Convolutional Network. Results show that the proposed approach significantly outperforms the state-of-the-art BERT model with orders of magnitude fewer number of parameters.

This paper presents our submission to task 5 ( Multimedia Automatic Misogyny Identification) of the SemEval 2022 competition. The purpose of the task is to identify given memes as misogynistic or not and further label the type of misogyny involved. In this paper, we present our approach based on language processing tools. We embed meme texts using GloVe embedding and classify misogyny using BERT model. Our model obtains an F1-score of 66.24% and 63.5% in misogyny classification and misogyny labels, respectively.

This paper describes and examines different systems to address Task 6 of SemEval-2021: Detection of Persuasion Techniques In Texts And Images, Subtask 1. The task aims to build a model for identifying rhetorical and psycho- logical techniques (such as causal oversimplification, name-calling, smear) in the textual content of a meme which is often used in a disinformation campaign to influence the users. The paper provides an extensive comparison among various machine learning systems as a solution to the task. We elaborate on the pre-processing of the text data in favor of the task and present ways to overcome the class imbalance. The results show that fine-tuning a RoBERTa model gave the best results with an F1-Micro score of 0.51 on the development set.

This paper describes the system submitted to SemEval-2021 Task-7 for all four subtasks. Two subtasks focus on detecting humor and offense from the text (binary classification). On the other hand, the other two subtasks predict humor and offense ratings of the text (linear regression). In this paper, we present two different types of fine-tuning methods by using linear layers and bi-LSTM layers on top of the pre-trained BERT model. Results show that our system is able to outperform baseline models by a significant margin. We report F1 scores of 0.90 for the first subtask and 0.53 for the third subtask, while we report an RMSE of 0.57 and 0.58 for the second and fourth subtasks, respectively.

Getting manually labeled data in each domain is always an expensive and a time consuming task. Cross-domain sentiment analysis has emerged as a demanding concept where a labeled source domain facilitates a sentiment classifier for an unlabeled target domain. However, polarity orientation (positive or negative) and the significance of a word to express an opinion often differ from one domain to another domain. Owing to these differences, cross-domain sentiment classification is still a challenging task. In this paper, we propose that words that do not change their polarity and significance represent the transferable (usable) information across domains for cross-domain sentiment classification. We present a novel approach based on χ2 test and cosine-similarity between context vector of words to identify polarity preserving significant words across domains. Furthermore, we show that a weighted ensemble of the classifiers enhances the cross-domain classification performance.

Identification of intensity ordering among polar (positive or negative) words which have the same semantics can lead to a fine-grained sentiment analysis. For example, ‘master’, ‘seasoned’ and ‘familiar’ point to different intensity levels, though they all convey the same meaning (semantics), i.e., expertise: having a good knowledge of. In this paper, we propose a semi-supervised technique that uses sentiment bearing word embeddings to produce a continuous ranking among adjectives that share common semantics. Our system demonstrates a strong Spearman’s rank correlation of 0.83 with the gold standard ranking. We show that sentiment bearing word embeddings facilitate a more accurate intensity ranking system than other standard word embeddings (word2vec and GloVe). Word2vec is the state-of-the-art for intensity ordering task.

For fine-grained sentiment analysis, we need to go beyond zero-one polarity and find a way to compare adjectives (synonyms) that share the same sense. Choice of a word from a set of synonyms, provides a way to select the exact polarity-intensity. For example, choosing to describe a person as benevolent rather than kind1 changes the intensity of the expression. In this paper, we present a sense based lexical resource, where synonyms are assigned intensity levels, viz., high, medium and low. We show that the measure P (s|w) (probability of a sense s given the word w) can derive the intensity of a word within the sense. We observe a statistically significant positive correlation between P(s|w) and intensity of synonyms for three languages, viz., English, Marathi and Hindi. The average correlation scores are 0.47 for English, 0.56 for Marathi and 0.58 for Hindi.