Readability: Difference between revisions

From Clarin K-Centre
Jump to navigation Jump to search
No edit summary
Line 14: Line 14:


==Machine learning==
==Machine learning==
The demo will process your text by deriving various text characteristics or features and predict a readability score using supervised machine learning techniques.


* The demo will process your text by deriving various text characteristics or features and predict a readability score using supervised machine learning techniques.
*[https://lt3.ugent.be/resources/machine-learning-readability/ Information]
*[https://lt3.ugent.be/resources/machine-learning-readability/ Information]
* De Clercq, Orphée and Véronique Hoste. 2016. All Mixed Up? Finding the Optimal Feature Set for General Readability Prediction and its Application to English and Dutch. Computational Linguistics, Association for Computational Linguistics, 42(3):457-490.
* De Clercq, Orphée and Véronique Hoste. 2016. All Mixed Up? Finding the Optimal Feature Set for General Readability Prediction and its Application to English and Dutch. Computational Linguistics, Association for Computational Linguistics, 42(3):457-490.
Line 21: Line 21:


==T-scan ==
==T-scan ==
* T-Scan is an analysis tool for Dutch text, mainly focusing on text complexity.
T-Scan is an analysis tool for Dutch text, mainly focusing on text complexity.
* [https://github.com/UUDigitalHumanitieslab/tscan/raw/master/docs/tscanhandleiding.pdf Manual (in Dutch)]
* [https://github.com/UUDigitalHumanitieslab/tscan/raw/master/docs/tscanhandleiding.pdf Manual (in Dutch)]
* [https://tscan.hum.uu.nl/tscan/ Tool]
* [https://tscan.hum.uu.nl/tscan/ Tool]

Revision as of 09:59, 16 February 2024

Assessing readability

How to assess readability? In this demo you can assess the readability of ten texts by comparing them with each other. The idea is that you assign an absolute score - ranging from 0 (easy) to 100 (difficult) - to each text and motivate your score using the free text field.

  • Information
  • De Clercq, Orphée and Véronique Hoste. 2016. All Mixed Up? Finding the Optimal Feature Set for General Readability Prediction and its Application to English and Dutch. Computational Linguistics, Association for Computational Linguistics, 42(3):457-490.
  • Demo

Classical formulas

In this demo, you can enter a Dutch text of maximum 1,000 characters. The text is then analyzed: various text characteristics (word length, sentence length, TTR, ...) are outputted and different scores calculated based on classical readability formulas. In a next phase, the text is also analyzed with a syntactic parser offering insights into the grammatical complexity of the text.

  • Information
  • De Clercq, Orphée and Véronique Hoste. 2016. All Mixed Up? Finding the Optimal Feature Set for General Readability Prediction and its Application to English and Dutch. Computational Linguistics, Association for Computational Linguistics, 42(3):457-490.
  • Demo

Machine learning

The demo will process your text by deriving various text characteristics or features and predict a readability score using supervised machine learning techniques.

  • Information
  • De Clercq, Orphée and Véronique Hoste. 2016. All Mixed Up? Finding the Optimal Feature Set for General Readability Prediction and its Application to English and Dutch. Computational Linguistics, Association for Computational Linguistics, 42(3):457-490.
  • Demo

T-scan

T-Scan is an analysis tool for Dutch text, mainly focusing on text complexity.