Proceedings of the Second Workshop on Natural Language Processing for Internet Freedom: Censorship, Disinformation, and Propaganda

24 papers
Assessing Post Deletion in Sina Weibo: Multi-modal Classification of Hot Topics (D19-50)

Copied to clipboard

Challenge: Weibo monitors and deletes posts to conform to government requirements . a recent study found that sentiment is the only indicator of censorship that is consistent across topics .
Approach: They analyze a dataset of censored and uncensore censors on Weibo . they use deep learning, CNN localization, and NLP techniques to analyze the data .
Outcome: The proposed analysis of censored and uncensoreded posts in Weibo shows that sentiment is the only indicator of a topic's censorship . censors can remove posts that are considered sensitive in three hours on average .
Detecting context abusiveness using hierarchical deep learning (D19-50)

Copied to clipboard

Challenge: Abusive text is a serious problem in social media and causes many issues among users . a model that detects text abusiveness in context without explicit abusive words is challenging .
Approach: They propose to use an abusive lexicon to determine the existence of an abusive word in text . they combine local and global features to evaluate the model using benchmark data .
Outcome: The proposed model outperforms all previous models for detecting abusiveness in text without abusive words.
How Many Users Are Enough? Exploring Semi-Supervision and Stylometric Features to Uncover a Russian Troll Farm (D19-50)

Copied to clipboard

Challenge: Social media has been used by troll farms to promote political agendas . trolled farms employ people to provoke conflict via the use of inflammatory or provocative comments.
Approach: They analyze the use of self-supervision with less than 100 troll accounts as training data to determine whether a trolled account is labeled as a Russian trol farm.
Outcome: The proposed methods improve classification performance by nearly 4% F1 and use self-supervision with less than 100 troll accounts as training data.
Identifying Nuances in Fake News vs. Satire: Using Semantic and Linguistic Cues (D19-50)

Copied to clipboard

Challenge: a blurry line between fake news and protected-speech satire has been a struggle for social media platforms . purveyors of fake news have begun to masquerade as satirical sites to avoid being demoted .
Approach: They propose to automatically classify fake news versus satire based on language differences . they hypothesize that nuances could be identified using semantic and linguistic cues .
Outcome: The proposed method can identify nuances between fake news and satire based on language differences . the proposed method is compared to the language-based baseline and is highly scalable .
Calls to Action on Social Media: Detection, Social Impact, and Censorship Potential (D19-50)

Copied to clipboard

Challenge: Calls to action are effective means of mobilization in social networks, but their potential for censorship and predicting offline protest events has not yet been evaluated.
Approach: They examine the possibility of their automatic detection on historical data from the 2011-2013 protests in Bolotnaya, Russia.
Outcome: The political calls to action can be annotated and detected with relatively high accuracy and have a moderate positive correlation with actual rally attendance.
Mapping (Dis-)Information Flow about the MH17 Plane Crash (D19-50)

Copied to clipboard

Challenge: Digital media enables fast sharing of information, but also disinformation . studies on the spread of disinformation on social media focused on small, manually annotated datasets or used proxys for data annotation.
Approach: They propose to use text classifiers to label Twitter content related to the MH17 crash to improve annotation accuracy.
Outcome: The proposed classifier improves over a hashtag-based baseline, but still remains a challenge in labelling pro-Russian and pro-Ukrainian content with high precision.
Generating Sentential Arguments from Diverse Perspectives on Controversial Topic (D19-50)

Copied to clipboard

Challenge: ArgDiver model generates high-quality sentential arguments from multiple perspectives . retrieval-based systems do not have sufficient flexibility for input with missing keywords or topics unseen .
Approach: They propose a neural method to generate sentential arguments from multiple perspectives . their model generates high-quality sentential argument, but shows higher diversity .
Outcome: The proposed model generates high-quality sentential arguments from multiple perspectives . it shows that it can provide diverse perspectives on a controversial topic .
Rumor Detection on Social Media: Datasets, Methods and Opportunities (D19-50)

Copied to clipboard

Challenge: Social media platforms are used for information gathering, but they also lead to the spreading of rumors and fake news.
Approach: This paper presents a comprehensive list of datasets used for rumor detection . it also reviews the important studies based on what types of information they exploit .
Outcome: This paper presents an overview of the recent studies in the rumor detection field . it provides a comprehensive list of datasets used for rumour detection .
Unraveling the Search Space of Abusive Language in Wikipedia with Dynamic Lexicon Acquisition (D19-50)

Copied to clipboard

Challenge: Existing methods to detect abusive language only train one classifier for the whole variety of offending . a new method can support a moderator with explicit unraveled explanations for why something was flagged as abusive .
Approach: a new method is proposed to distinguish explicitly abusive cases from the more "shadowed" ones . the researchers extend a lexicon of abusive terms to include new obfuscations of abusive words .
Outcome: a new method can distinguish explicitly abusive cases from the more "shadowed" ones . the method can support a moderator with explicit unraveled explanations for why something was flagged as abusive .
CAUnLP at NLP4IF 2019 Shared Task: Context-Dependent BERT for Sentence-Level Propaganda Detection (D19-50)

Copied to clipboard

Challenge: Sentence-level and fragment-level propaganda detection tasks are more challenging compared to document-level detection.
Approach: They propose to use context-dependent input pairs to fine-tune the pretrained propaganda detection BERT to better utilize document information.
Outcome: The proposed system can detect propaganda on document-level, sentence-level and fragment-level.
Fine-Grained Propaganda Detection with Fine-Tuned BERT (D19-50)

Copied to clipboard

Challenge: The goal of the Fragment Level Classification task is to detect and classify textual segments that correspond to one of the 18 given propaganda techniques in a news articles dataset.
Approach: They propose a model that performs word-level classification using a pre-trained language model to detect and classify propaganda fragments in a news article dataset.
Outcome: The proposed model performs word-level classification using a popular pre-trained language model.
Neural Architectures for Fine-Grained Propaganda Detection in News (D19-50)

Copied to clipboard

Challenge: MIC-CIS is a fine grained propaganda detection system . previous work focused on document level, labeling articles as propaganda .
Approach: They propose to use different neural architectures to jointly perform propaganda detection tasks . they also investigate different ensemble schemes such as majority-voting, relax-vote, etc.
Outcome: The proposed system performs sentences and fragment level propaganda detection tasks.
Fine-Tuned Neural Models for Propaganda Detection at the Sentence and Fragment levels (D19-50)

Copied to clipboard

Challenge: The system was evaluated on a unified development set without distributing the gold labels.
Approach: They propose to use fine-grained propaganda detection to build models that can explain why an article is propagandistic.
Outcome: The proposed model performed on all eighteen propaganda techniques in the corpus of the shared task.
Divisive Language and Propaganda Detection using Multi-head Attention Transformers with Deep Learning BERT-based Language Models for Binary Classification (D19-50)

Copied to clipboard

Challenge: a team of researchers used a pre-trained BERT language model to train propaganda . the model was based on a cloze comprehension test to answer a question about influence operations .
Approach: team used a BERT language model that was pre-trained on Wikipedia and BookCorpus . they used cloze comprehension tests to train the model to answer a propaganda question .
Outcome: The proposed model was trained on Wikipedia and BookCorpus to answer propaganda questions . the team used a neural network that was pre-trained on the Wikipedia and bookCorpus corpus .
On Sentence Representations for Propaganda Detection: From Handcrafted Features to Word Embeddings (D19-50)

Copied to clipboard

Challenge: The rise of fake and hyperpartisan news on social media and online news outlets calls for improved automatic detection of propaganda in texts.
Approach: They propose to use handcrafted features and learn dense semantic representations to detect propaganda in sentence-level and with random undersampling of the majority class (non-propaganda)
Outcome: The proposed system achieves a ranking of 10 among 25 participants, with 59.5 F1-score.
JUSTDeep at NLP4IF 2019 Task 1: Propaganda Detection using Ensemble Deep Learning Models (D19-50)

Copied to clipboard

Challenge: Detecting fake news is not well established yet, but it can be classified under several labels: false, biased, or framed to mislead the readers.
Approach: They propose a deep learning model using BiLSTM, XGBoost, and BERT to detect propaganda using a corpus from a challenge.
Outcome: The proposed model outperforms the baseline model on a dataset from the challenge NLP4IF 2019 .
Detection of Propaganda Using Logistic Regression (D19-50)

Copied to clipboard

Challenge: Various propaganda techniques are used to manipulate peoples perspectives to foster a predetermined agenda.
Approach: They propose a Logistic Regression-based tool that automatically classifies whether a sentence is propagandistic or not.
Outcome: The proposed tool outperforms the baseline on linguistic and semantic features.
Cost-Sensitive BERT for Generalisable Sentence Classification on Imbalanced Data (D19-50)

Copied to clipboard

Challenge: Popular NLP tasks such as sentiment analysis and event extraction from social media are examples of imbalanced classification problems.
Approach: They propose a method to generalise on dissimilar training and test data using a measure of similarity between datasets.
Outcome: The proposed method achieves the second highest score on sentence-level propaganda classification.
Understanding BERT performance in propaganda analysis (D19-50)

Copied to clipboard

Challenge: Despite the challenging nature of the shared task, our pretrained BERT model scored 0.62 F1 on the test set and ranked third among 25 teams who participated in the contest.
Approach: They propose to use a dataset to fine-tune a model for propaganda analysis at sentence level to determine whether a text is 'propaganda' and to examine false-positive cases.
Outcome: The proposed model scored 0.62 F1 on the test set and ranked third among 25 teams who participated in the shared task.
Pretrained Ensemble Learning for Fine-Grained Propaganda Detection (D19-50)

Copied to clipboard

Challenge: Propaganda detection is a reallife challenge that can affect how people understand news .
Approach: They propose to use a manually annotated dataset to tackle the propaganda detection on sentence level classification task of NLP4IF 2019 workshop co-located with EMNLP-IJCNLP 2019 conference.
Outcome: The proposed model is ranked in the first place with 68.8312 F1-score on the development dataset and in the sixth place with 61.3990 F1 score on the testing dataset.
NSIT@NLP4IF-2019: Propaganda Detection from News Articles using Transfer Learning (D19-50)

Copied to clipboard

Challenge: In this paper, we describe our approach and system description for NLP4IF 2019 Workshop: Shared Task on Fine-Grained Propaganda Detection.
Approach: They propose to use document Embeddings and LSTM to detect whether a sentence contains a propagandistic agenda.
Outcome: The proposed approach ranked 21st in the NLP4IF 2019 Workshop: Shared Task on Fine-Grained Propaganda Detection.
Sentence-Level Propaganda Detection in News Articles with Transfer Learning and BERT-BiLSTM-Capsule Model (D19-50)

Copied to clipboard

Challenge: a new task is needed to detect propaganda in news articles . the need for communication has increased in online social media platforms . a proposed solution to the problem of sentence-level propaganda classification is ranked 12th .
Approach: They propose to build a binary classifier able to provide corresponding propaganda labels . their solution ranks 12th among 26 teams in the NLP4IF-2019 Shared Task SLC .
Outcome: The proposed model outperforms baseline approach and the winning system on a similar task.
Synthetic Propaganda Embeddings To Train A Linear Projection (D19-50)

Copied to clipboard

Challenge: Using contextualized token embeddings, we can extract features of propaganda from contextualized embeddnings without fine-tuning the large parameters of the base model.
Approach: They propose a method for detecting fine-grained categories of propaganda in text by generating synthetically generated embeddings from pre-trained language models.
Outcome: The proposed method is used in the first shared task in fine-grained propaganda detection at NLP4IF as Team Stalin.
Findings of the NLP4IF-2019 Shared Task on Fine-Grained Propaganda Detection (D19-50)

Copied to clipboard

Challenge: A shared task on fine-grained propaganda detection was organized at EMNLP-IJCNLP 2019 . 12 systems submitted systems for the FLC task, 25 for the SLC task, and 14 teams submitted a system description paper .
Approach: They present a task on fine-grained propaganda detection as part of the NLP4IF workshop at EMNLP-IJCNLP 2019 . they used a corpus of news articles annotated with an inventory of propagandist techniques at the fragment level to determine the propaganda technique used in each fragment .
Outcome: The shared task on fine-grained propaganda detection was organized at the EMNLP-IJCNLP 2019 . 12 systems submitted for the FLC task, 25 for the SLC task, and 14 submitted a system description paper .

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations