-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathindex.html
160 lines (157 loc) · 14.1 KB
/
index.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="utf-8">
<meta name="viewport" content="width=device-width, initial-scale=1, shrink-to-fit=no">
<title>Combating Online Scientific Misinformation</title>
<meta name="keywords" content="Misinformation, Scientific Misinformation, COVID-19 Misinformation, Fact-Checking, Computational Journalism, Scientific Journalism, Science Communication, Scientific News, SciClops, SciLens, SciLander" lang="en">
<meta name="description" content="Combating Online Scientific Misinformation">
<meta name="author" content="Panayiotis Smeros">
<link rel="shortcut icon" type="image/png" href="assets/img/sci.png">
<!-- Font Awesome icons (free version)-->
<script src="https://use.fontawesome.com/releases/v5.15.1/js/all.js" crossorigin="anonymous"></script>
<!-- Google fonts-->
<link href="https://fonts.googleapis.com/css?family=Saira+Extra+Condensed:500,700" rel="stylesheet" type="text/css">
<link href="https://fonts.googleapis.com/css?family=Muli:400,400i,800,800i" rel="stylesheet" type="text/css">
<!-- Core theme CSS (includes Bootstrap)-->
<link href="css/styles.css" rel="stylesheet">
</head>
<body id="page-top">
<!-- Navigation-->
<nav class="navbar navbar-expand-lg navbar-dark bg-primary fixed-top" id="sideNav">
<a class="navbar-brand js-scroll-trigger" href="#page-top">
<span class="d-block d-lg-none">Combating Online Scientific Misinformation</span>
<span class="d-none d-lg-block"><img class="img-fluid img-profile rounded-circle mx-auto mb-2" src="assets/img/sci.png" alt=""></span>
</a>
<button class="navbar-toggler" type="button" data-toggle="collapse" data-target="#navbarSupportedContent" aria-controls="navbarSupportedContent" aria-expanded="false" aria-label="Toggle navigation"><span class="navbar-toggler-icon"></span></button>
<div class="collapse navbar-collapse" id="navbarSupportedContent">
<ul class="navbar-nav">
<li class="nav-item"><a class="nav-link js-scroll-trigger" href="#about">About</a></li>
<li class="nav-item"><a class="nav-link js-scroll-trigger" href="#SciClops">SciClops</a></li>
<li class="nav-item"><a class="nav-link js-scroll-trigger" href="#SciLens">SciLens</a></li>
<li class="nav-item"><a class="nav-link js-scroll-trigger" href="#SciLander">SciLander</a></li>
<li class="nav-item"><a class="nav-link js-scroll-trigger" href="#details">Details</a></li>
</ul>
</div>
</nav>
<!-- Page Content-->
<div class="container-fluid p-0">
<!-- About-->
<section class="resume-section" id="about">
<div class="resume-section-content">
<h2 class="mb-0"><span class="text-primary">Combating Online Scientific Misinformation</span></h2>
<div class="subheading mb-0"><a href="https://panayiotis.smeros.info">Panayiotis Smeros</a>, <a href="https://chato.cl">Carlos Castillo</a>, <a href="https://people.epfl.ch/karl.aberer/?lang=en">Karl Aberer</a></div>
<div class="subheading mb-5">PhD Thesis at <a href="https://www.epfl.ch/en/">EPFL</a></div>
<p class="lead mb-3" style="text-align: justify;">
The drastic shift towards <b>digital communication</b> in our mediasphere has caused a profound change in the production and consumption of information, which in turn has substantial implications on the <b>social and political landscape</b>.
Misinformation, as a side effect of mass information diffusion, has become a fundamental problem for <b>governments, platforms, and the general public</b> in light of critical events such as <b>elections, pandemics, and wars</b>.
In this thesis, we focus on the problem of online <b>scientific misinformation</b>, proposing methods that tackle the problem at different granularity levels: at the level of claims, articles, and sources.
</p>
<div class="social-icons text-center">
<a class="social-icon" href="mailto:panayiotis.smeros@alumni.epfl.ch"><i class="fa fa-envelope"></i></a>
<a class="social-icon" href="https://twitter.com/psmeros"><i class="fab fa-twitter"></i></a>
<a class="social-icon" href="https://linkedin.com/in/psmeros"><i class="fab fa-linkedin"></i></a>
<a class="social-icon" href="https://github.com/psmeros"><i class="fab fa-github"></i></a>
</div>
</div>
</section>
<hr class="m-0">
<section class="resume-section" id="SciClops">
<div class="resume-section-content">
<h2 class="mb-5">SciClops</h2>
<p class="lead mb-3" style="text-align: justify;">
For combating claim-based scientific misinformation, we introduce <b>SciClops</b>, a method for detecting and contextualizing scientific claims for assisting manual fact-checking.
Our method involves three steps: (1) extracting scientific claims using a <b>domain-specific, fine-tuned transformer model</b>, (2) clustering similar claims together with <b>related scientific literature</b> using a method that exploits their content and the connections among them, and (3) highlighting check-worthy claims broadcasted by <b>popular yet unreliable sources</b>.
Our experiments show that SciClops effectively assists non-expert fact-checkers in verifying complex scientific claims, facilitating them to <b>outperform commercial fact-checking systems</b>.
</p>
<p class="lead mb-0 text-center">
<img src="assets/sci-img/SciClops_Overview.png" alt="SciClops Overview" style="height:10em">
</p>
<p class="lead mb-0 text-center">
SciClops Overview
</p>
</div>
</section>
<hr class="m-0">
<section class="resume-section" id="SciLens">
<div class="resume-section-content">
<h2 class="mb-5">SciLens</h2>
<p class="lead mb-3" style="text-align: justify;">
For combating article-based scientific misinformation, we introduce <b>SciLens</b>, a method for evaluating the quality of scientific news articles.
Our method involves a series of quality indicators for news articles that derive from: (1) their content, including the use of <b>attributed quotes</b>, (2) their scientific context, including their <b>semantic similarity and web proximity to the scientific literature</b>, and (3) their social context, including their <b>social media reach and stance</b>.
Our experiments show that <b>these indicators help non-experts</b> evaluate the quality of articles more accurately compared to non-experts that do not have access to these indicators.
Moreover, SciLens can also produce completely <b>automated quality scores for articles</b>, which agree more with expert evaluators than manual evaluations done by non-experts.
</p>
<p class="lead mb-0 text-center">
<img src="assets/sci-img/SciLens_Overview.png" alt="SciLens Overview" style="height:10em">
</p>
<p class="lead mb-0 text-center">
SciLens Overview
</p>
</div>
</section>
<hr class="m-0">
<section class="resume-section" id="SciLander">
<div class="resume-section-content">
<h2 class="mb-5">SciLander</h2>
<p class="lead mb-3" style="text-align: justify;">
For combating source-based scientific misinformation, we introduce <b>SciLander</b>, a method for learning representations of news sources reporting on scientific topics.
Our method involves heterogeneous source indicators that capture: (1) the <b>copying</b> of news stories between sources, (2) the <b>semantic shift</b> of terms across sources, (3) the usage of <b>jargon</b>, and (4) the <b>stance</b> towards specific citations.
SciLander uses these indicators as signals of source agreement to train <b>unsupervised source embeddings</b>.
Our experiments show that the learned source representations outperform state-of-the-art baselines on the task of <b>news veracity classification</b> while encoding information about the <b>reliability, political leaning, and partisanship bias</b> of these sources.
</p>
<p class="lead mb-0 text-center">
<img src="assets/sci-img/SciLander_Overview.png" alt="SciLander Overview" style="height:10em">
</p>
<p class="lead mb-0 text-center">
SciLander Overview
</p>
</div>
</section>
<hr class="m-0">
<!-- Details-->
<section class="resume-section" id="details">
<div class="resume-section-content">
<h2 class="mb-5">Details</h2>
<div class="subheading mb-0">Publications</div>
<ul class="fa-ul mb-3">
<li>[<b>ICWSM'23</b>] M. Gruppi, P. Smeros, S. Adali, C. Castillo, K. Aberer. <em>SciLander: Mapping the Scientific News Landscape.</em> [<a href="files/icwsm2023.pdf">pdf</a>, <a href="files/icwsm2023.bib">bib</a>]</li>
<li>[<b>CIKM'21</b>] P. Smeros, C. Castillo, K. Aberer. <em>SciClops: Detecting and Contextualizing Scientific Claims for Assisting Manual Fact-Checking.</em> [<a href="files/cikm2021.pdf">pdf</a>, <a href="files/cikm2021.pptx">slides</a>, <a href="files/cikm2021.bib">bib</a>]</li>
<li>[<b>VLDB'20</b>] A. Romanou, P. Smeros, C. Castillo, K. Aberer. <em>SciLens News Platform: A System for Real-Time Evaluation of News Articles.</em> [<a href="files/vldb2020.pdf">pdf</a>, <a href="files/vldb2020.bib">bib</a>]</li>
<li>[<b>WWW'19</b>] P. Smeros, C. Castillo, K. Aberer. <em>SciLens: Evaluating the Quality of Scientific News Articles Using Social Media and Scientific Literature Indicators.</em> [<a href="files/www2019.pdf">pdf</a>, <a href="files/www2019.pptx">slides</a>, <a href="files/www2019.bib">bib</a>]</li>
</ul>
<div class="subheading mb-0">System</div>
<ul class="fa-ul mb-3">
<li><a href="https://newsteller.lsir.ch">NewsTeller</a>: Real-Time News Analytics Platform</li>
</ul>
<div class="subheading mb-0">Code, Models, Datasets</div>
<ul class="fa-ul mb-3">
<li>Code Release for <a href="https://github.com/psmeros/SciClops">SciClops</a>, <a href="https://github.com/psmeros/SciLens">SciLens</a>, and <a href="https://github.com/psmeros/SciLander">SciLander</a></li>
<li><a href="https://huggingface.co/psmeros/NewsBERT">NewsBERT</a> and <a href="https://huggingface.co/psmeros/SciNewsBERT">SciNewsBERT</a> models, and <a href="data/SciClops_validation_set.csv">validation set</a> for <b>scientific claim extraction</b></li>
<li>Anonymized evaluations of <b>scientific claims</b> by <a href="data/SciClops_expert_evaluations.csv">experts</a>, <a href="data/SciClops_non_expert_evaluations.csv">non-experts</a>, and <a href="data/SciClops_commercial_systems_evaluations.csv">commercial systems</a></li>
<li>Anonymized evaluations of <b>scientific news articles</b> by <a href="data/SciLens_expert_evaluations.csv">experts</a> and <a href="data/SciLens_non_expert_evaluations.csv">non-experts</a></li>
<li>Anonymized <b>social media postings</b>, <b>news articles</b>, and <b>scientific papers</b> in <a href="data/diffusion_graph.tsv.bz2">graph format</a></li>
<li>Lists of <a href="data/academic_repositories.csv">Academic Repositories</a> and <a href="data/institution_domains.csv">Institution Domains</a></li>
<li>Vocabularies for <a href="data/Health_and_Nutrition_Vobaculary.txt">Health and Nutrition</a>, <a href="data/Science_in_News_Vocabulary.txt">Science in News</a>, and <a href="data/COVID-19_Vobaculary.txt">COVID-19</a></li>
</ul>
<div class="subheading mb-0">Support</div>
<ul class="fa-ul mb-3">
<li><a href="https://www.epfl.ch/research/open-science/champions/">Open Science Fund</a></li>
<li><a href="https://www.satw.ch/en/">Swiss Academy of Engineering Sciences</a></li>
</ul>
</div>
</section>
<hr class="m-0">
</div>
<footer class="footer">
<div class="text-center"><small class="copyright">Last update 10/2022. Template designed by <a href="https://startbootstrap.com/theme/resume" target="_blank">Start Bootstrap</a>.</small></div>
</footer>
<!-- Bootstrap core JS-->
<script src="https://cdnjs.cloudflare.com/ajax/libs/jquery/3.5.1/jquery.min.js"></script>
<script src="https://cdn.jsdelivr.net/npm/bootstrap@4.5.3/dist/js/bootstrap.bundle.min.js"></script>
<!-- Third party plugin JS-->
<script src="https://cdnjs.cloudflare.com/ajax/libs/jquery-easing/1.4.1/jquery.easing.min.js"></script>
<!-- Core theme JS-->
<script src="js/scripts.js"></script>
</body>
</html>