IITD at the WANLP 2022 Shared Task: Multilingual Multi-Granularity Network for Propaganda Detection
This work addresses propaganda detection in Arabic social media, but it is incremental as it builds on existing multilingual models without major breakthroughs.
The authors tackled propaganda detection in Arabic tweets by developing a system for multi-label classification and sequence tagging, achieving second place in both subtasks among 14 and 3 participants, respectively.
We present our system for the two subtasks of the shared task on propaganda detection in Arabic, part of WANLP'2022. Subtask 1 is a multi-label classification problem to find the propaganda techniques used in a given tweet. Our system for this task uses XLM-R to predict probabilities for the target tweet to use each of the techniques. In addition to finding the techniques, Subtask 2 further asks to identify the textual span for each instance of each technique that is present in the tweet; the task can be modeled as a sequence tagging problem. We use a multi-granularity network with mBERT encoder for Subtask 2. Overall, our system ranks second for both subtasks (out of 14 and 3 participants, respectively). Our empirical analysis show that it does not help to use a much larger English corpus annotated with propaganda techniques, regardless of whether used in English or after translation to Arabic.