Deep Active Learning with Simulated Rationales for Text Classification

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Neural networks have become a preferred tool for text classification tasks, demonstrating state of the art performances when trained on a large set of labeled data. However, in an early active learning setup, the scarcity of the ground-truth labels available severely penalizes the generalization capability of the neural network. In order to overcome such limitations, in this paper, we introduce a new learning strategy, which consist of inserting in the early stages of the learning process some additional, local and salient knowledge, presented under the form of simulated, human like rationales. We show how such knowledge can be automatically extracted from documents by analyzing the class activation maps of a convolutional neural network. The experimental results obtained demonstrate that the exploitation of such rationales permits to significantly speed-up the learning process, with a spectacular increase of the accuracy rates, starting from a very reduced number of documents (10–20).

Original languageEnglish
Title of host publicationPattern Recognition and Artificial Intelligence - International Conference, ICPRAI 2020, Proceedings
EditorsYue Lu, Nicole Vincent, Pong Chi Yuen, Wei-Shi Zheng, Farida Cheriet, Ching Y. Suen
PublisherSpringer Science and Business Media Deutschland GmbH
Pages363-379
Number of pages17
ISBN (Print)9783030598297
DOIs
Publication statusPublished - 1 Jan 2020
Event2nd International Conference on Pattern Recognition and Artificial Intelligence, ICPRAI 2020 - Zhongshan, China
Duration: 19 Oct 202023 Oct 2020

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume12068 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference2nd International Conference on Pattern Recognition and Artificial Intelligence, ICPRAI 2020
Country/TerritoryChina
CityZhongshan
Period19/10/2023/10/20

Keywords

  • Active learning
  • Class activation maps
  • Deep neural networks
  • Rationales
  • Text classification

Fingerprint

Dive into the research topics of 'Deep Active Learning with Simulated Rationales for Text Classification'. Together they form a unique fingerprint.

Cite this