Summary of A Multi-task Text Classification Pipeline with Natural Language Explanations: a User-centric Evaluation in Sentiment Analysis and Offensive Language Identification in Greek Tweets, by Nikolaos Mylonas et al.
A Multi-Task Text Classification Pipeline with Natural Language Explanations: A User-Centric Evaluation in Sentiment Analysis and Offensive Language Identification in Greek Tweets
by Nikolaos Mylonas, Nikolaos Stylianou, Theodora Tsikrika, Stefanos Vrochidis, Ioannis Kompatsiaris
First submitted to arxiv on: 14 Oct 2024
Categories
- Main: Computation and Language (cs.CL)
- Secondary: Machine Learning (cs.LG)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary This paper proposes a novel pipeline for generating natural language explanations in text classification tasks, particularly sentiment analysis and offensive language identification. The pipeline consists of two models: a classifier for labelling texts and an explanation generator that provides the reasoning behind predictions. Unlike existing interpretability techniques, this approach produces explanations that can be easily understood by non-expert users. The proposed pipeline is evaluated through a user study using three metrics on Greek tweets, achieving promising results for both datasets. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper develops a new way to explain how machines make decisions about text. Instead of just saying what’s important or giving rules, this method provides explanations in simple language that anyone can understand. The system is made up of two parts: one that guesses the right label for a piece of text and another that explains why it chose that label. This approach can be used for many different tasks, like figuring out if someone is being mean online or what people think about something. It was tested on Greek tweets and worked well for both sentiment analysis and detecting offensive language. |
Keywords
» Artificial intelligence » Text classification
