Social media has become incredibly popular these days for communicating with friends and for sharing opinions. According to current statistics, almost 2.22 billion people use social media in 2016, which is roughly one third of the world population and three times of the entire population in Europe. In social media people share their likes, dislikes, opinions, interests, etc. so it is possible to know about a person’s thoughts about a specific topic from the shared data in social media. Since, twitter is one of the most popular social media in the world; it is a very good source for opinion mining and sentiment analysis about different topics. In this research, SVM with different kernel functions and Adaboost are experimented using CPD and Chi-square feature extraction techniques to explore the best sentiment classification model. The reported average accuracy of Adaboost for Chi-square and CPD are 70.2% and 66.9%. The SVM radial basis kernel and polynomial kernel with Chi-square n-grams reported average accuracy of 73.73% and 68.67% respectively. Among the performed experimentation, SVM sigmoid kernel with Chi-square n-grams provided the maximum accuracy that is 74.4%.

Existing Natural Language Inference (NLI) datasets, while being instrumental in the advancement of Natural Language Understanding (NLU) research, are not related to scientific text. In this paper, we introduce SCINLI, a large dataset for NLI that captures the formality in scientific text and contains 107, 412 sentence pairs extracted from scholarly papers on NLP and computational linguistics. Given that the text used in scientific literature differs vastly from the text used in everyday language both in terms of vocabulary and sentence structure, our dataset is well suited to serve as a benchmark for the evaluation of scientific NLU models. Our experiments show that SCINLI is harder to classify than the existing NLI datasets. Our best performing model with XLNet achieves a Macro F1 score of only 78.18% and an accuracy of 78.23% showing that there is substantial room for improvement.

No abstract

Natural Language Inference (NLI) or Recognizing Textual Entailment (RTE) aims at predicting the relation between a pair of sentences (premise and hypothesis) as entailment, contradiction or semantic independence. Although deep learning models have shown promising performance for NLI in recent years, they rely on large scale expensive human-annotated datasets. Semi-supervised learning (SSL) is a popular technique for reducing the reliance on human annotation by leveraging unlabeled data for training. However, despite its substantial success on single sentence classification tasks where the challenge in making use of unlabeled data is to assign "good enough" pseudo-labels, for NLI tasks, the nature of unlabeled data is more complex: one of the sentences in the pair (usually the hypothesis) along with the class label are missing from the data and require human annotations, which makes SSL for NLI more challenging. In this paper, we propose a novel way to incorporate unlabeled data in SSL for NLI where we use a conditional language model, BART to generate the hypotheses for the unlabeled sentences (used as premises). Our experiments show that our SSL framework successfully exploits unlabeled data and substantially improves the performance of four NLI datasets in low-resource settings. We release our code at: https:// github.com/msadat3/SSL_for_NLI.

No abstract