In the 4th ICWSM 2023 data challenge, we invite papers that model and contribute to understanding the temporal and social dynamics of the social tasks in the provided datasets, or identify other important related dimensions to study the temporal effect on those datasets.

We welcome submissions on various topics that address the temporal shift of data. The data challenge includes three tracks, 1) Time-aware models and social trends, 2 ) Temporal dataset, and 3) Non- archival track.


Track 1: Time-Aware Models and social trends ( archival)


Task1 : Leveraging the evolution in time for specific task.

We encourage submissions that characterize the incorporation of temporal data in different social-based tasks, both at data and model level. Tasks may include, but are not limited to, modeling temporal characteristics of hate speech, using different versions of edited Wikipedia articles, enhanced tone detection, the effectiveness of temporal data in detecting the veracity of rumors, and stance detection dynamics. Submitted papers may focus on evaluating models' performance when considering the time variable, studying the evolution of specific phenomena, examining the distribution shift between the training data and live update data, or focusing on specific concept shifts.


Task2: Studying trends and social change.

We especially welcome contributions that examine the transitions in community or content based on social data (i.e., temporal networks or content) to analyze some social phenomena. For example, studying context or static embedding to analyze content across time and communities where the main goal could be to study changes in descriptions of genders and ethnic groups, representation of people using contextualized semantics, analyzing factors that cause the formation and persistence of trends, the dynamics of sentiment and topics, or the manifestation of different social phenomena across communities.

Track 2: Temporal Dataset (archival)

Participants are also welcome to submit their own temporal dataset and will be part of the full proceedings. The work should be in compliance with The FAIR Data Principles,

Datasets and metadata must be published using a dataset sharing service (e.g. Zenodo , datorium , dataverse , or any other dataset sharing services that index your dataset and metadata and increase the re-findability of the data) that provides a DOI for the dataset, which should be included in the dataset paper submission. Ethical considerations must be discussed and datasets need to comply with the platform API regulations and follow the general principle that guarantees the transparency of datasets, such as datasheets for datasets.


Track 3: Non-Archival

The non-archival track seeks recently accepted/published work as well as work-in-progress. It does not need to be anonymized and will not go through the review process. The submission should clearly indicate the original venue and will be accepted.




Further detail about the Data Challenge

https://sites.google.com/view/icwsm2023datachallenge/home

ICWSM 2023 Data challenge