Semeval-2016 Task 4 : Sentiment Analysis in Twitter

By G. Briones, Kasun Amarasinghe, Bridget T. McInnes. 2016

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Explaining Recurrent Neural Network Predictions in Sentiment Analysis (2017, background): “Recently, a trend emerged for tackling these challenges via deep learning models such as convolutional and recurrent neural networks, as observed e.g. on the SemEval-2016 Task for Sentiment Analysis in Twitter (Nakov et al., 2016).”
Using Convolutional Neural Networks to Classify Hate-Speech (2017, methodology): “…and sentiment analysis, although most of the best performing systems in SemEval 2016 (the International Workshop on Semantic Evaluation; Task 4: Sentiment Analysis in Twitter) actually utilised convolutional neural networks or combinations of CNNs and other approaches (Nakov et al., 2016).”
DynaSent: A Dynamic Benchmark for Sentiment Analysis (2020, background): “SemEval 2016 and 2017 (Nakov et al., 2016; Rosenthal et al., 2017) offered Twitter-based sentiment datasets that continue to be used widely despite their comparatively small size ( ≈ 10K examples).”
SemEval-2018 Task 3: Irony Detection in English Tweets (2018, background): “They enriched the training set with 1,000 humorous tweets from SemEval-2017 Task 6 (Potash et al., 2017) and another 1,000 tweets with positive polarity from SemEval-2016 Task 4 (Nakov et al., 2016), resulting in a training corpus of 5,834 tweets.”
Hate Me, Hate Me Not: Hate Speech Detection on Facebook (2017, methodology): “This type of neural network has been recently tested on sentiment analysis tasks [33, 37], reaching outperforming classification performance [29], with even a 3-4 points improvement with respect to commonly used learning algorithms.”
Findings of the 2017 DiscoMT Shared Task on Cross-lingual Pronoun Prediction (2017, methodology): “While in 2015 we used macro-averaged F 1 as an ofﬁcial evaluation measure, this year we followed the setup of 2016, where we switched to macro-averaged recall , which was also recently adopted by some other competitions, e.g., by SemEval-2016/2017 Task 4 (Nakov et al., 2016; Rosenthal et al., 2017).”
Fidelity-Weighted Learning (2017, methodology): “To make your results comparable to the ofﬁcial runs on SemEval we us SemEval-14 (1,853 tweets) and SemEval-15 (2,390 tweets) as test sets (Rosenthal et al., 2015; Nakov et al., 2016).”
Using millions of emoji occurrences to learn any-domain representations for detecting sentiment, emotion and sarcasm (2017, background): “The third dataset is from SemEval 2016 Task4A (Nakov et al., 2016).”
A Survey on Text Classification: From Traditional to Deep Learning (2020, background): “The SA datasets include Movie Review (MR) [257] [226], Stanford Sentiment Treebank (SST) [227], Multi-Perspective Question Answering (MPQA) [258] [229], IMDB [230], Yelp [231], Amazon Reviews (AM) [93], NLP&CC 2013 [111], Subj [250], CR [251], SS-Twitter [259], SS-Youtube [259], SE1604 [260] and so ”
SemEval-2017 Task 4: Sentiment Analysis in Twitter (2017, methodology): “For evaluating binary quantification, we keep the Kullback-Leibler Divergence (KLD) measure used in 2016 along with additive smoothing (Nakov et al., 2016a; Forman, 2005).”
Biased Embeddings from Wild Data: Measuring, Understanding and Removing (2018, methodology): “In the domain of text, many modern approaches often begin by embedding the input text data into an embedding space that is used as the ﬁrst layer in a subsequent deep network [4], [14].”
SemEval-2019 Task 8: Fact Checking in Community Question Answering Forums (2019, methodology): “We have described SemEval 2019 Task 8 on Fact Checking in Community Question Answering Forums.”
Masking as an Efficient Alternative to Finetuning for Pretrained Language Models (2020, background): “…datasets that have publicly available test sets: the 6-class question classiﬁcation dataset TREC-6 (Voorhees and Tice, 2000), the 4-class news classi-ﬁcation dataset AG News (AG) (Zhang et al., 2015), and the binary Twitter sentiment classiﬁcation task SemEval-2016 4B (SEM) (Nakov et al., 2016).”
A Survey on Text Classification: From Shallow to Deep Learning (2020, unknown): “F 5 93 3,650,000 [5] [46] [131] [190] SA DSTC 4 89 - 30,000 [9] [191] DAC MRDA 5 - 62,000 [9] [192] DAC SwDA 43 - 1,022,000 [9] [193] DAC RCV1 103 240 807,595 [117] [123] [68] [134] [139] [194] NC RCV1-V2 103 124 804,414 [74] [139] [195] NC NLP&CC 2013 2 - 115,606 [125] [126] SA SS-Twitter 2 - 2,113”
From classification to quantification in tweet sentiment analysis (2016, background): “Tweet quantiﬁcation is one of the subjects of the recent SemEval Task 4 ‘‘Sentiment Analysis in Twitter’’ shared task (Nakov et al. 2016), where tweets are labelled according to the sentiment they convey towards a certain topic.”
DataStories at SemEval-2017 Task 4: Deep LSTM with Attention for Message-level and Topic-based Sentiment Analysis (2017, background): “Whereas, arti-ﬁcial neural networks (Deriu et al., 2016; Rouvier and Favre, 2016) which perform feature learning, last year (Nakov et al., 2016) achieved very good results, outperforming the competition.”
Multitask Learning for Fine-Grained Twitter Sentiment Analysis (2017, methodology): “Support Vector Machines (SVMs) are maximum margin classi cation algorithms that have been shown to achieve competitive performance in several text classi cation problems [19].”
Projecting Embeddings for Domain Adaption: Joint Modeling of Sentiment Analysis in Diverse Domains (2018, methodology): “In order to evaluate how well domain adaptation techniques perform on large domain gaps, we also use the message polarity classiﬁcation corpora provided by the organizers of SemEval 2013 and 2016 (Nakov et al., 2013; Nakov et al., 2016).”