Tweets are made of 4 dimensions you can analyze. We have specific analyses or “Recipes” for each of the dimensions: CONTENT COMPONENT + SOCIAL STRUCTURE + EXTERNAL INFORMATION LINKS + KEYWORDS
For the tweets , these text analyses can be centered in the content of the original tweets. Tweets will be clustered by topics and info on number of retweets, likes, sentiment, organizations mentioned and hashtags will be obtained.
You have a context, or a Twitter list, or a groups of twitter users and:
1.You want to know what a community is talking about, in general, which narratives they generate
2.You can find which accounts talk more about a specific word you are looking for (eg. blockchain, IoT, the name of your brand)
3.You can track your competitors and analyse their hashtags, people they mention, and the impact of their tweets
You want to track the content that a specific twitter account generates, and see what narratives do they talk about, and which have more repercussion.
You have a hashtag you want to track, and you want to see which groups of tweets by which accounts had more RTs and what narratives these groups generated.
You want to track all mentions to your twitter alias or the name of your brand. See the different narratives around it, and how those messages propagate.
Combine the people and the content use cases, restricting the analysis of keywords to a specific group of people, or vice versa.
1.Extract the data - Tractor configuration:
Select the Use case you want to work on, and download the associated data from Tractor (see tractor support for search queries).
2.Recipe selection in Graphext: Tweets content topics.
Connections: content similarity Text processing and analysis of tweets on the studied narratives.
Identify narratives and their significant terms, temporal patterns and narratives that get more attention and impact.
We know that data isn't always clean and simple.
Have a look through these topics if you can't see what you are looking for.