TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

Better Sentiment Analysis with Sentiment.ai

2 点作者 shrewdcomputer大约 3 年前

1 comment

uniqueuid大约 3 年前
Mandatory warning: Whatever you&#x27;re measuring with sentiment detection, it&#x27;s probably not sentiment, and it&#x27;s extremely inconsistent across implementations and humans.<p>This wonderful paper shows how unreliable sentiment detection is: <a href="https:&#x2F;&#x2F;www.tandfonline.com&#x2F;doi&#x2F;pdf&#x2F;10.1080&#x2F;19312458.2020.1869198" rel="nofollow">https:&#x2F;&#x2F;www.tandfonline.com&#x2F;doi&#x2F;pdf&#x2F;10.1080&#x2F;19312458.2020.18...</a><p>Summary: (1) The best performance is still attained with trained human or crowd coding; (2) None of the used dictionaries come close to acceptable levels of validity; and (3) machine learning, especially deep learning, substantially outperforms dictionary-based methods but falls short of human performance