This repository contains scripts and experiments associated with the RoMemes corpus
RoMEMES is a dataset of Romanian language memes, collected from public social media platforms. The dataset was manually annotated with: associated text in Romanian language, image complexity, sentiment, political content. In addition, the dataset contains associated metadata and the text part was automatically annotated in the RELATE platform with part-of-speech tags, lemmas, and dependency parsing.
The dataset can be downloaded here: https://zenodo.org/doi/10.5281/zenodo.13120215
Citation for the dataset: Păiș, V., Niță, S., Jerpelea, A., Pană, L., & Curea, E. (2024). RoMEMES [Data set]. Zenodo. https://doi.org/10.5281/zenodo.13120216