Impact of Quality Of Images On Users Behavior On Social Media

The last decade has witnessed one of the fastest growth in Web 2.0 technology with the diversity of channels, structure and content of user generated content. Such data can be textual, visual or multimedia. This created valuable source of data for users, professionals and research community. This paper explores potential correlation that may exist between images and the associated textual content. Especially, we explore the influence of the image quality on the nature of the users’ content. For this purpose, Brisque image quality index has been performed. The quantification between the image quality and the users’ textual generated content is evaluated using both the number of comments generated by users and overall sentiment score associated to the given image. The results of such analysis confirms the hypothesis of positive correlation between the quality of posted image and the associated user’s content in the above sense. A new tourism dataset Tourism481 constructed from IMGUR web service was also introduced and made public.

Authors:
Bounab Yazid, Oussalah Mourad, Beddiar Djamila Romaissa

Publication type:
A4 Article in conference proceedings

Place of publication:
2020 Tenth International Conference on Image Processing Theory, Tools and Applications (IPTA)

Keywords:
image captioning, image description, sentiments, Social Media

Published:

Full citation:
Y. Bounab, M. Oussalah and D. R. Beddiar, “Impact of Quality Of Images On Users Behavior On Social Media,” 2020 Tenth International Conference on Image Processing Theory, Tools and Applications (IPTA), Paris, 2020, pp. 1-6, doi: 10.1109/IPTA50016.2020.9286466

DOI:
https://doi.org/10.1109/IPTA50016.2020.9286466

Read the publication here:
http://urn.fi/urn:nbn:fi-fe202101262656