Information Propagation by Composited Labels in Natural Language Processing
Computation and Language
2022-05-24 v1 Artificial Intelligence
Machine Learning
Abstract
In natural language processing (NLP), labeling on regions of text, such as words, sentences and paragraphs, is a basic task. In this paper, label is defined as map between mention of entity in a region on text and context of entity in a broader region on text containing the mention. This definition naturally introduces linkage of entities induced from inclusion relation of regions, and connected entities form a graph representing information flow defined by map. It also enables calculation of information loss through map using entropy, and entropy lost is regarded as distance between two entities over a path on graph.
Cite
@article{arxiv.2205.11509,
title = {Information Propagation by Composited Labels in Natural Language Processing},
author = {Takeshi Inagaki},
journal= {arXiv preprint arXiv:2205.11509},
year = {2022}
}