联邦神经主题模型

Dec, 2022

Federated Neural Topic Models

Lorena Calvo-Bartolomé, Jerónimo Arenas-García

TL;DR本文探讨了用于主题建模的联邦学习方法，以解决跨数据源分析时的隐私问题，并提出一种基于最先进的神经网络主题模型实现的联邦实现方案，保护各节点的隐私。实验结果表明，与中心化方法相比，该方法在跨节点文档主题多样性和构建联合模型方面更具优势。

Abstract

Over the last years, topic modeling has emerged as a powerful technique for organizing and summarizing big collections of documents or searching for particular patterns in them. However, privacy concerns arise wh