IIIT Hyderabad Publications |
|||||||||
|
Neural Approaches Towards Computational JournalismAuthor: Dhruv Khattar Date: 2019-03-05 Report no: IIIT/TH/2019/9 Advisor:Vasudeva Varma,Manish Gupta AbstractGone are the days when individuals aimlessly pursued news from one source. Presently, they need to peruse articles and stories from an assortment of sources to see every one of its subtleties. News aggregators become an integral factor here. They gather news articles from an assortment of sources and present it to the client in one single area. With news originating from different streams, it turns out to be exceptionally cumbersome for a client to choose articles of her decision from a rundown of exhibited articles which relates to an assortment of subjects. In this thesis, we mainly focus on three major aspects of Computational Journalism which could help news aggregators to increase viewership as well as their credibility. We start with news recommendation, which is also one of the most crucial things for a news aggregator. Users, these days don’t have time to browse through hundreds of news articles and find the relevant ones. They want a recommendation system which could take into account their reading history and recommend relevant articles. We then look at a new problem which is used by news agencies to lure users into clicking at headlines and subsequently increase their viewership, which is formally known as Clickbait. Finally, we try to detect Fake News which could help news agencies in building trust and increase their credibility. We commence our research work with the problem of News Recommendation. Popular methods like collaborative filtering and content-based filtering have their own disadvantages. The former method requires a considerable amount of user data before making predictions, while the latter, suffers from over-specialization. In this work, we address both of these issues by coming up with a hybrid approach based on neural networks for news recommendation. The hybrid approach incorporates for both (1) user-item interaction and (2) content-information of the articles read by the user in the past. We first come up with an article-embedding based profile for the user. We then use this user profile with adequate positive and negative samples in order to train the neural network-based model. We then move on to various deep learning methods to tackle the same problem. We propose a Re-current Attention Recommendation Engine (RARE) which consists of two components and utilizes the distributed representations of news articles. The first component is used to model the user’s sequential behaviour of news reading in order to understand her general interests, i.e., to get a summary of her interests. The second component utilizes an article level attention mechanism to understand her specific interests. We feed the information obtained from both the components to a Siamese Network in order to make predictions which pertain to the user’s generic as well as specific interests. Building upon our previous model, we propose a Hybrid Recurrent Attention Machine (HRAM).HRAM consists of two components. The first component utilizes a neural network for matrix factoriza-tion. While in the second component, we first learn the distributed representation of each news article.We then use the historical data of the user in a sequential manner and feed it to an attention-based recurrent layer. Finally, we concatenate the outputs from both these components and use it to make predictions. In this way, we harness the information present in the user reading history and boost it with the information available through collaborative filtering for providing better news recommendations. We then experiment with non-recurrent neural networks for the same problem. We propose a simple yet effective architecture which utilizes a 3D Convolutional Neural Network which takes the word em-beddings of the articles present in the user history as its input. Using such a method endows the model with the capability to automatically learn spatial (features of a particular article) as well as temporal features (features across articles read by a user) which signify the interest of the user. At test time, we use this in combination with a 2D Convolutional Neural Network for recommending articles to users. We carry out extensive experiments to establish the effectiveness of our methods. We also perform experiments to prove the efficacy of our model on solving the item cold-start cases as well as making effective recommendations for users who have had very little interaction with articles. We then look at another aspect of Computational Journalism, which is Clickbait Detection. Online media outlets, in a bid to expand their reach and subsequently increase revenue, have begun adopting clickbait techniques to lure readers to click on articles. We propose a novel approach considering all information found in a social media post. We train a bidirectional LSTM with an attention mechanism to learn the extent to which a word contributes to the post’s clickbait score in a differential manner. We also employ a Siamese net to capture the similarity between source and target information. We learn image embeddings from large amounts of data using Convolutional Neural Networks to add another layer of complexity to our model. Finally, we concatenate the outputs from the three separate components, serving it as input to a fully connected layer. We conduct experiments over a test corpus of 19538 social media posts bettering the previous state-of-the-art. Finally, we look at Multimodal Fake News Detection. Given the prominence of microblogging net-works as a source of news for most individuals, fake news now spreads at a faster pace and has a more profound impact than ever before. This makes detection of fake news an extremely important challenge. A shortcoming of the current approaches for the detection of fake news is their inability to learn a shared representation of multimodal (textual + visual) information. We propose an end-to-end network, Multimodal Variational Autoencoder (MVAE), which uses a bimodal variational autoencoder coupled with a binary classifier for the task of fake news detection. The model consists of three main components, an encoder, a decoder and a fake news detector module. The variational autoencoder is capable of learning probabilistic latent variable models by optimizing a bound on the marginal likelihood of the observed data. The fake news detector then utilizes the multimodal representations obtained from the bimodal variational autoencoder to classify posts as fake or not. We conduct extensive experiments on two standard fake news datasets collected from popular microblogging websites: Weibo and Twitter. The experimental results show that across the two datasets, our model outperforms state-of-the-art methods. Full thesis: pdf Centre for Search and Information Extraction Lab |
||||||||
Copyright © 2009 - IIIT Hyderabad. All Rights Reserved. |