Analysis on news content in the Distroid Database
Early work on analyzing content to find some insights in the Distroid Database.
The most frequent content type was news, with 282 entries. The least frequent was analytics dashboard with 3 entries.
You can find more information in the table and graphic below.
For Idea Machines (IM), Web3 was the most common IM. You can find more information in the table and graphic below.
Platform Cooperativism, Web3
Tools for Thought
Tools for Thought, Web3
Platform Cooperativism, Web3, Responsible Technology
Open Science, Platform Cooperativism
Tools for Thought, Open Science
Tools for Thought, Web3, Open Science
I focused on news articles to start the analysis of the database for issues 28-38.
In total, there were 282 news items in the database. I reduced the news items to 274 after removing N/A or NaN values from the title column.
At most, an author or publication appeared twice in the dataset.
I then used spaCy with NLTK to conduct sentiment analysis on titles and text. The sentiment was usually positive for an article’s text.
This might be reflective of the high-spirited, optimistic, and forward-looking nature of content written about the frontier.
The sentiment for an article’s title was usually neutral.
I created a graph visualization of the relationship between authors, publications, and news articles.
I still need to work on my visualization skills (especially with graphs). So stay tuned for better visualizations in the next update.
I also created an interactive graph visualization with pyvis.
The color scheme for the interactive pyvis visualization is:
red = Author,
purple = Publication, and
blue = Article.
You can find the visualization on GitHub as an HTML file. Load the HTML file on your browser and you can interact with the graph visualization.
I also made wordcloud visualizations for news titles, text, keywords, and summaries.