Cross-sectional Vector Search
2023-04-08
We've been doing keyword exact match cross search of our collection for a long time.
2022's pLinkSuggest allows "search for a sentence by sentence". The advantage at this time was that by taking the "common string of ambiguous matches", it became clear "why the sentence was suggested".
Vector search alone is the same as mere recommendation
I don't see why it matters.
'Are the hits relevant to the current theme?' If so, in what ways? If so, in what ways?" You can go so far as to have them explain
2023-09-15
"Are the hits relevant to the current theme? Or not? If so, in what ways are they relevant?" to the point where they are asked to explain
You noticed this back in April.
As of September, we think this is really important.
The term "recommendation with context" was coined.
---
This page is auto-translated from /nishio/蔵書横断ベクトル検索 using DeepL. If you looks something interesting but the auto-translated English is not good enough to understand it, feel free to let me know at @nishio_en. I'm very happy to spread my thought to non-Japanese readers.