What Opensolr Adds to Every Document
Before a page or a document lands in your Opensolr Index, Opensolr reads it and adds what search needs: its language, its tone, how much real content it has, its price and category, its video details, its meaning and the text inside its pictures. You write no code for any of it.
These additions are what the visitor filters, the ranking, the hybrid search, the out-of-the-box chat bot and the answers are built on.
Crawled pages and pushed documents
Pages Opensolr crawls and documents you push with Data Ingestion get the same core additions. A few come only from reading a live web page.
| What is added | Crawled pages | Pushed documents |
|---|---|---|
| Language and locale | Yes | Yes, unless you send your own |
| Sentiment | Yes | Yes |
| Meaning vector | Yes, with AI in the plan | Yes, with AI in the plan |
| Text in scans and pictures | PDFs found on your site | Files sent by URL with rtf:true, or uploaded |
| Pictures described | Pictures with no text in PDFs | Same, in the files you send |
| Content quality score | Yes | No, counted as average |
| Prices, categories, breadcrumbs | Yes, from structured data | Send them as your own fields |
| Video details | Yes, from video markup | Send them as your own fields |
The meaning vector and the picture descriptions use the AI of your plan: what site search uses from your plan. The fields each addition writes are listed in Fields Opensolr adds to each document.
Where you see them
- Visitor filters on the hosted search page: language, locale, price, video duration. See visitor filters and facet filters.
- Ranking: the quality score feeds the Content Quality Boost.
- Search by meaning: the meaning vector feeds hybrid search.
- The chat bot quotes prices and reads PDF pages, with the text found in their pictures: Chat Bot.
What every page gets
- What every page gets
- Language and locale
- Sentiment
- Content quality score
- Prices and categories
- Video details