Fork me on GitHub
#xtdb
<
2023-04-15
>
sb15:04:12

I hope someone has experience with this, as I would like to use Lucene’s similarity function. I’m trying to find similar text, like “hello” for example, using the cosine algorithm. How can this be achieved? may I need to setup index for this, but how to do?

refset21:04:52

Hey @U2T2ZEVPC I worked a lot on this module and wrote the extension_test namespace, but unfortunately I don't have understanding or experience with configuring the Lucene similarity settings - would be happy to bounce ideas around though or help you decipher the Lucene docs. Certainly you have a lot of freedom to configure, extend, or even fork the module to fit your needs 🙂

👍 2
👀 2
sb01:04:27

That is great!! I found in the Lucene docs, but not so clear how to do. I’m back with this in week with my questions/ after tests. Thanks in advance 👍

🙏 2