With the exponential growth of users and user-generated content present on online social networks, fake news and its detection have become a major problem. Through these, smear campaigns can be generated, aimed for example at trying to change the political orientation of some people. Twitter has become one of the main spreaders of fake news in the network. Therefore, in this paper, we present a solution based on Text Mining that tries to find which text patterns are related to tweets that refer to fake news and which patterns in the tweets are related to true news. To test and validate the results, the system faces a pre-labelled dataset of fake and real tweets during the U.S. presidential election in 2016. In terms of results interesting patterns are obtained that relate the size and subtle changes of the real news to create fake news. Finally, different ways to visualize the results are provided.

/pdf/mining-text-patterns-over-fake-and-real-tweets-53a3vegzfh.pdf

Mining Text Patterns over Fake and Real Tweets

AI in Pathology: What could possibly go wrong?

Social networks have taken an irreplaceable role in our lives. They are used daily by millions of people to communicate and inform themselves. This success has also led to a lot of irrelevant content and even misinformation on social media. In this paper, we propose a user-centred framework to reduce the amount of irrelevant content in social networks to support further stages of data mining processes. The system also helps in the reduction of misinformation in social networks, since it selects credible and reputable users. The system is based on the belief that if a user is credible then their content will be credible. Our proposal uses word embeddings in a first stage, to create a set of interesting users according to their expertise. After that, in a later stage, it employs social network metrics to further narrow down the relevant users according to their credibility in the network. To validate the framework, it has been tested with two real Big Data problems on Twitter. One related to COVID-19 tweets and the other to last United States elections on 3rd November. Both are problems in which finding relevant content may be difficult due to the large amount of data published during the last years. The proposed framework, called NOFACE, reduces the number of irrelevant users posting about the topic, taking only those that have a higher credibility, and thus giving interesting information about the selected topic. This entails a reduction of irrelevant information, mitigating therefore the presence of misinformation on a posterior data mining method application, improving the obtained results, as it is illustrated in the mentioned two topics using clustering, association rules and LDA techniques. 

https://digibug.ugr.es/bitstream/10481/76708/1/1-s2.0-S0957417422012684-main.pdf

NOFACE: A new framework for irrelevant content filtering in social media according to credibility and expertise

This work presents the application of several Deep Learning techniques for Natural Language Processing to the classification of tweets into containing fake news or not. To validate our approach, we use an open-access dataset containing annotated tweets related to the 2016 US elections. From our experiments, we can confirm that Deep Learning techniques are indeed able to identify tweets containing fake news, and that LSTMs with pre-computed embeddings is the best performing among the tested techniques (validation AUC = 0.70), particularly in avoiding misclassification of the minority class.

Deep learning for fake news classification

<h3>Introduction</h3> CT angiography (CTA), the modality of choice for detection of intracranial aneurysms (IA), is widely available at all care levels. Implementing deep learning algorithms (DLA) in the radiologists’ workflow could result in higher detection rates, particularly in smaller centers. We present results of multicenter pre-clinical experience with a new IA detection DLA. <h3>Methods</h3> A DLA was developed at Site 1 to predict the existence and locations of IAs &gt;2 mm in 3D head CTA volumes. It localizes IAs using an improved 3D UNet to regress the bounding boxes of suspicious locations. IA presence is confirmed by a multi-resolution 3D DenseNet at each suspicious location. The model was trained on 436 studies and tested on 696 studies from site 1. To validate robustness against protocol variability, it was retrospectively tested on 337 volumes (149 positives, 188 negatives) from 4 additional sites with various types of scanners. A neuroradiologist/trained fellow from each site annotated bounding boxes around the IAs with 3DSlicer. The model’s performance was assessed by comparing its predicted aneurysm locations to the annotated boxes. <h3>Results</h3> The model achieved a lesion-level sensitivity of 0.72 (95% CI: 0.66 - 0.75) with 1.4 false positives per volume (FPPV, 95% CI: 1.22 - 1.57). Sensitivity ranged 0.84 (1.72 FPPV) - 0.67 (1.40 FPPV) amongst 5 sites. Performance variation was secondary to testing on images from scanners the model had not been trained on (unseen). Removing unseen scanners, the model achieved sensitivity of 0.77 (95% CI: 0.68 - 0.84) with 1.51 FPPV (95% CI: 1.24 - 1.82). Patient-level area under curves (AUCs) on the external sites were comparable to that of site 1 dataset, which was 0.88 (95% CI: 0.84 - 0.92). The model performed better in the detection of internal carotid and anterior communicating artery IAs &gt; 3 mm. <h3>Conclusions</h3> This DLA for IA detection showed robust performance in detecting IAs &gt; 3 mm. DLAs may reliably help detection of IAs on CTA, particularly in the community practices less familiar with intracranial vascular imaging. Different scanners can lead to performance variability. This can be avoided by training with data from different scanners. <h3>Disclosures</h3> <b>D. Wu:</b> 1; C; National Institute of Health. <b>E. Orru’:</b> None. <b>S. Ferraciolli:</b> None. <b>A. Ashok:</b> 1; C; NIHR Academic Clinical Fellowship. <b>A. Sorby-Adams:</b> None. <b>M. Guindy:</b> None. <b>D. Montes:</b> None. <b>A. Medina:</b> None. <b>G. Bianco:</b> None. <b>J. Kaggie:</b> 1; C; NIHR Cambridge BRC. <b>D. Rabina:</b> None. <b>O. Ashush:</b> None. <b>X. Han:</b> None. <b>Y. Baror:</b> None. <b>M. Patel:</b> 4; C; Rhino HealthTech Inc. <b>I. Dayan:</b> 4; C; Rhino HealthTech Inc.. <b>J. Romero:</b> None. <b>R. González:</b> 1; C; National Institute of Health. 2; C; MR Access, Inc.. 4; C; MR Access, Inc.. <b>C. Wald:</b> None. <b>F. Kitamura:</b> 2; C; MD.ai, GE Healthcare. <b>P. Kuriki:</b> None. <b>T. Matys:</b> 1; C; NIHR Cambridge BRC. 2; C; NHS. <b>Q. Li:</b> 1; C; National Institute of Health. 

/pdf/e-107-multi-center-study-of-a-deep-learning-model-for-eqbin8mv.pdf

Juan Gómez Romero

Papers

Deep learning for fake news classification

E-107 Multi-center study of a deep learning model for intracranial aneurysm detection in computed tomography angiography