We used a multiview-learning for regression that exploits word embedding models and linguistic resource (word length, frequency lists, lexical databases composed of school dictionaries). We use it to infer four psycholinguistic properties of words:
- Concreteness;
- Subjective Frequency;
- Imageability;
- AoA
This work is adaption to Inferring Psycholinguistic Properties of Words (Paetzold and Specia; 2016) for portuguese language.
Draft version A Lightweight Regression Method to Infer Psycholinguistic Properties for Brazilian Portuguese
Final version will appear in Proceedings of 20th international conference on Text, Speech and Dialogue.
- scikit-learn (0.18.1)
- scipy (0.18.1)
- nltk (3.2.1)
- keras (2.0)
For more information, you can contact me via sborgesleandro@gmail.com or leandrobs@usp.br.