Virality factors on Linkedin
The project consists in understanding the origin of virality on Linkedin through a semantic approach applied to a dataset of more than 4 million posts.
Several statistical analyses have already been carried out, revealing the influence of 3 specific factors, such as emojis and the length of the post.
In the next steps of the project we will carry out a fine-grained manual semantic analysis of the most viral posts in order to identify the factors that are likely to explain the engagement of Internet users.
The role of emojis has been studied after a through investigation of their distribution in a dataset of 4.6m posts. We have used a language detection algorithm and further analyzed the popularity of emojis per language.
Since the use of a language doesn’t necessarily correspond to the country of the user, we repeated the same analysis for the countries with significant samples.
These prelimiary analyses led to further discoveries on the impact of emojis and virality.