A Web-Based Language Analysis Tool!
Linguine Text Analysis User Test #1
Notes
- For any Stanford CoreNLP analyses, any ‘Tokenize’ Preprocessing Options will not effect the outcome of analyses. They can be selected without error but will not change the result.
Setup
- Complete the pre-testing survey here
- Download and unzip the Fresh.txt zip file
- Login to Linguine using your DCE (RIT Username. Ex: abc1234@rit.edu) account: http://nlp.rit.edu/linguine/
- Please open the Linguine app in a new window in order to follow directions from this webpage in the tool
Activity
- Add the Fresh.txt file as a corpus to the Linguine tool. You can name it anything you like.
- Using the Fresh.txt file as the corpus, perform a Named Entity Recognition Analysis. Choose any preprocessing options you would like.
- How many NUMBER entities are recognized?
- Hint: switch from the ‘Visualization’ tab to the ‘Default View’ tab and use the search feature to look for ‘NUMBER’.
- Using the Fresh.txt file as the corpus, perform a Term Frequency Analysis. For preprocessing options feel free to select any that you like.
- What is the most frequently used word?
- How many unique words (word types) are there?
Debriefing
- Complete the post-testing survey here