Would you like to take the introduction tour to the demo? (Recommended for first time users)
You have concluded the introduction tour! You are now ready to explore some datasets on your own.
Click Ok to go back and select a dataset.
SenSE is a tool that facilitates the analysis and exploration of Semantic Shift (also called Language Change or Semantic Change).
Semantic shift is the process through which words in a given language are subject to changes in meaning. Such changes may consist of a word acquiring a new sense, driven by cultural transformations. The acquisition of a new sense by a word may be followed by a total loss of its original sense.
For example, the word awful in English in the year 1800 was used to describe something impressive (full of awe). Eventually, awful a synonym for something unpleasant and it is no longer used in its original sense.
Another example is that of the word plane which before the invention of the aircraft related to a surface and in the 21st century it acquired new meaning despite still being used to designate a surface.
In this demonstration, you will be able to explore semantic shift in datasets from different time periods, cultures, communities, and languages. You will be able to inspect words that are semantically shifted, exploring difference in contexts, similar words, and sentence examples that are semantically distinct from each other.
Choose the dataset to work with. The datasets consist of pairs of corpora from different domains or periods.
English texts from different time periods.
Collections of German text from two different historical periods.
Texts in Latin from different historical periods.
Two corpora of the Swedish language from different historical periods.
ArXiv papers from Artificial Intelligence and Classical Physics.
List of the most semantically shifted words. You can select a word from this list to further inspect them.
Enter a word to query its neighbors in each corpus. These results show the difference in contexts for the same word in each corpus.
Examples of sentences that exhibit the distinct semantics between a word in each input corpus.