In recent years various studies have revealed what influence the spread of misinformation on social media can have on people’s beliefs. Most famous is the influence misinformation had during the U.S. elections of 2016 and the Brexit referendum. After these events some claimed that a post-truth era has started which refers to the disappearance of shared objective standards for truth. To contribute to the fight against misinformation ICMCIS challenges you to develop misinformation detection algorithms that can automatically classify Tweets (Twitter posts) as related to misinformation or trusted information, focusing on both (self-derived) linguistic and network features that indicate misinformation.
The data challenge is hosted using Kaggle’s InClass competition tool. To participate in this challenge, and find more information about the provided data, you should visit the following website: https://www.kaggle.com/t/26a170db3c0c486594e9a013b690ad10.
By participating in this challenge you are asked to submit a short paper about your findings. These papers will be published as conference proceedings and participants are invited to present them during the 21st ICMCIS in Amsterdam. More information about the submission requirements of this paper will be provided when the competition has started.
Important to know: Due to privacy constraints of Twitter only the identifiers of Tweets are provided and should be used to extract Twitter data. Since Twitter is actively removing suspicious Tweets/Twitter accounts it may occur that during this challenge some Tweets will become unavailable. To make sure all participant will use the same data a final dataset will be released on February 1st in which all unavailable Tweets are removed. All participants are obligated to do a submission on the final data set to compete in this challenge. More rules and regulations can be found here: https://www.kaggle.com/c/icmcis2020/rules.
Start challenge: 1 November 2019
Release Final dataset: 1 February 2020
Deadline Kaggle submissions: 1 March 2020
Deadline short paper: 1 April 2020