Multiple Instance Learning Networks for Fine-Grained Sentiment Analysis

Stefanos Angelidis, Mirella Lapata
11/27/2017
cs.CLcs.IRcs.LG

Official code: stangelid/milnet-sent (35 stars, last commit 2018-11-26).

Abstract

We consider the task of fine-grained sentiment analysis from the perspective of multiple instance learning (MIL). Our neural model is trained on document sentiment labels, and learns to predict the sentiment of text segments, i.e. sentences or elementary discourse units (EDUs), without segment-level supervision. We introduce an attention-based polarity scoring method for identifying positive and negative text snippets and a new dataset which we call SPOT (as shorthand for Segment-level POlariTy annotations) for evaluating MIL-style sentiment models like ours. Experimental results demonstrate superior performance against multiple baselines, whereas a judgement elicitation study shows that EDU-level opinion extraction produces more informative summaries than sentence-based alternatives.

View on arXivView PDF

Code Implementations(1)

stangelid/milnet-sentOfficial100%

Multiple Instance Learning Networks for Fine-Grained Sentiment Analysis

351Nov 21, 20177 years ago

Similar Papers

Aspect-Based Sentiment Analysis using Local Context Focus Mechanism with DeBERTa

87% match2022

USA: Universal Sentiment Analysis Model & Construction of Japanese Sentiment Text Classification and Part of Speech Dataset

86% match2023

Utilizing BERT for Aspect-Based Sentiment Analysis via Constructing Auxiliary Sentence

86% match1 repo5192019

iACOS: Advancing Implicit Sentiment Extraction with Informative and Adaptive Negative Examples

86% match2023

Transformation Networks for Target-Oriented Sentiment Classification

85% match1 repo1392018

Leveraging ChatGPT As Text Annotation Tool For Sentiment Analysis

85% match2023

Cite this paper

@article{angelidis2017multiple,
  title  = {Multiple Instance Learning Networks for Fine-Grained Sentiment Analysis},
  author = {Stefanos Angelidis and Mirella Lapata},
  year   = {2017},
  eprint = {1711.09645},
  archivePrefix = {arXiv},
  url    = {https://arxiv.org/abs/1711.09645}
}

Add a badge to your README

Shows whether official code is linked for this paper.

Code of Paper badge
[![Code of Paper](https://codeofpaper.com/badge/1711.09645.svg)](https://codeofpaper.com/paper/1711.09645)

Discussion