This repository is dedicated to usage of Lövheim cube of emotion model in tasks of sentiment analysis. Currenlty research is performed only for Russian language, but more languages are hoped to come.
data: A place for text data along with scrapers and loader.features: A place for pure feature functions (most linguistiscs related) to extract relevant params from texts.misc: Programm plumbings like typings and global constants.models: Place for storing trained models.projects: Main folder. Here we play and experiment with different approaches and strategies.tools: Tools for processing and cleaning data.
- Mapping Texts to Multidimensional Emotional Space: Challenges for Dataset Acquisition in Sentiment Analysis
- Automated Soundtrack Generation for Fiction Books Backed by Lövheim’s Cube Emotional Model
Feel free to contact on contribution and discussion: @AlexDel, xyz@verbalab.ru
The research is funded by Russian Foundation for Basic Research. Grant #19-012-00205