This interactive visualization shows you the differences based on the frequency and contextual relationship between two query words within the same dataset – this visualisation is especially useful for understanding and articulating the data landscape of close synonyms and antonyms, for example research vs experiment.
Two query words appear, one on each side of the visualisation, between them are words that have either a relationship or are used in context with the query words. The closer the word is to the center, the more it tends to be combined with both of the given words. The words that appear closest to the query words on the right / left shows the contextual closeness of the relationship with the query word.
The size of the circle and the text has a direct correlation to the frequently the word can be found in the dataset. You can see a detailed explanation by hovering over the words or by clicking on the circles surrounding the words.
To change the query, i.e. words, you can either click on a surrounding word or enter a new query word into search field. Simply click on a word on left or on right and a type field will appear, enabling you to enter new query words.
The differences were automatically generated based on algorithms that look for words which appear in contexts to the query words using corpus analysis tool SketchEngine.
The underlying data was taken from ACL Anthology Reference Corpus, a collection of scientific papers about Computational Linguistics.