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。

Show HN: Semantic Search on Wikipedia

1 点作者 jonas_kgomo大约 2 年前
This is a semantic search engine that understands the meaning of words and can return results that match the user&#x27;s intent, even if the words used in the search query are not an exact match for the text in the articles.<p>To achieve this, we used Cohere&#x27;s 100M embeddings on Wikipedia, which allowed us to map each article to a high-dimensional vector space [786] where similar articles are cosine close to each other.<p>We then built a simple web app that allows users to search for articles using natural language queries and returns the top results, along with a 2D visualization of the articles using UMAP.<p><a href="https:&#x2F;&#x2F;txt.cohere.com&#x2F;embedding-archives-wikipedia" rel="nofollow">https:&#x2F;&#x2F;txt.cohere.com&#x2F;embedding-archives-wikipedia</a>

暂无评论

暂无评论