Mini project-text mining
We explored both JSTOR TopicGraph and Voyant and compared their functionality below. Our testing article is The Public Sphere: An Encyclopedia Article by Jurgen Habermas. Before using those tools, we were expecting to extract the keywords or most frequently occuring words that imply overview of the article, and to understand network and logic among the extracted words.Both tools are pretty straightforward to use, but they display different features.
JSTOR TopicGraph allows us to compare the extracting information and articles horizontally. It extracts “topics” automatically and shows what topic are covered in this book. When clicking on graphic next to topic words, users are led directly to a page that discuss this topic. However, Topicgraph doesn’t display how the topics and their related terms are chosen. Users are not allowed to edit topics and related terms. The relationship and network among different topics are not analyzed either. http://voyant-tools.org/?corpus=eb5e4b4483871aa09c310d81c5bd51e1
VOYANT has more features in terms of analysing and visualizing the data. Besides the basic visualization of text frequency, there’s bubble diagram showing the relations among the key words; Filters are also provided for more accurate analysis;The number of text segment can also be customized, all these additional features provides more information and flexibility for the user to understand the text, making it more powerful than JSTOR in terms of data interpretation. https://labs.jstor.org/topicgraph/monograph/324348d0eab2692439be05e7217dae29