Buzz Tweet Classification Based on Text and Image Features of Tweets Using Multi-Task Learning

This study investigates social media trends and proposes a buzz tweet classification method to explore the factors causing the buzz phenomenon on Twitter. It is difficult to identify the causes of the buzz phenomenon based solely on texts posted on Twitter. It is expected that by limiting the tweets to those with attached images and using the characteristics of the images and the relationships between the text and images, a more detailed analysis than that of with text-only tweets can be conducted. Therefore, an analysis method was devised based on a multi-task neural network that uses both the features extracted from the image and text as input and the buzz class (buzz/non-buzz) and the number of “likes (favorites)” and “retweets (RTs)” as output. The predictions made using a single feature of the text and image were compared with the predictions using a combination of multiple features. The differences between buzz and non-buzz features were analyzed based on the cosine similarity between the text and the image. The buzz class was correctly identified with a correctness rate of approximately 80% for all combinations of image and text features, with the combination of BERT and VGG16 providing the highest correctness rate.

Collecting a Large Scale Dataset for Classifying Fake News Tweets Using Weak Supervision

The problem of automatic detection of fake news in social media, e.g., on Twitter, has recently drawn some attention. Although, from a technical perspective, it can be regarded as a straight-forward, binary classification problem, the major challenge is the collection of large enough training corpora, since manual annotation of tweets as fake or non-fake news is an expensive and tedious endeavor, and recent approaches utilizing distributional semantics require large training corpora. In this paper, we introduce an alternative approach for creating a large-scale dataset for tweet classification with minimal user intervention. The approach relies on weak supervision and automatically collects a large-scale, but very noisy, training dataset comprising hundreds of thousands of tweets. As a weak supervision signal, we label tweets by their source, i.e., trustworthy or untrustworthy source, and train a classifier on this dataset. We then use that classifier for a different classification target, i.e., the classification of fake and non-fake tweets. Although the labels are not accurate according to the new classification target (not all tweets by an untrustworthy source need to be fake news, and vice versa), we show that despite this unclean, inaccurate dataset, the results are comparable to those achieved using a manually labeled set of tweets. Moreover, we show that the combination of the large-scale noisy dataset with a human labeled one yields more advantageous results than either of the two alone.

Characterizing Social TV Activity Around Televised Events: A Joint Topic Model Approach

Viewers often use social media platforms like Twitter to express their views about televised programs and events like the presidential debate, the Oscars, and the State of the Union speech. Although this promises tremendous opportunities to analyze the feedback on a program or an event using viewer-generated content on social media, there are significant technical challenges to doing so. Specifically, given a televised event and related tweets about this event, we need methods to effectively align these tweets and the corresponding event. In turn, this will raise many questions, such as how to segment the event and how to classify a tweet based on whether it is generally about the entire event or specifically about one particular event segment. In this paper, we propose and develop a novel joint Bayesian model that aligns an event and its related tweets based on the influence of the event’s topics. Our model allows the automated event segmentation and tweet classification concurrently. We present an efficient inference method for this model and a comprehensive evaluation of its effectiveness compared with the state-of-the-art methods. We find that the topics, segments, and alignment provided by our model are significantly more accurate and robust.

Public Sentiments Analysis about Indonesian Social Insurance Administration Organization on Twitter

Insurance Administration Organization, which can be used by all people. However, this organization has received various criticisms from the public through social media, namely Twitter. This study aims to analyze public sentiment about the Indonesian Social Insurance Administration Organization on Twitter. The method used in this research is the Naive Bayes Classifier (NBC) method and uses the Support Vector Machine (SVM) method as a comparison. The amount of data used was 12,990 tweets with a data collection period from September 14, 2019 - February 18, 2020. The study compared the two classifier models built, namely the classifier model with two sentiment classes and four sentiment classes. The accuracy results show that the SVM method has a better accuracy value than the NBC method. SVM has an accuracy value of 63.60% and 82.77% for the two sentiment classes in the four sentiment classifier model. The tweet classification results show that the public's conversation about the Indonesian Social Insurance Administration Organization on Twitter has a negative polarity value tendency.