What to Ask Leina to Do with Dandelion API
Copy a prompt, add your details, and send it to Leina in your connected team chat.
Extract linked entities and turn their positions and references into a document index.
Extract Dandelion entities from this plain text [text]. Group the entities by relevance to [research question], include their source-text positions and links when available, and explain ambiguous references.
Use a semantic similarity result to guide a comparison of two supplied texts.
Compare Dandelion semantic similarity for [first text] and [second text]. Explain the score alongside shared ideas and differences in the supplied wording.
Get Started with Dandelion API
-
Connect Your Account
Connect with a Dandelion API token.
-
Try a Task
Tell Leina what you need in your team chat. Include the source, destination, and expected result.
Set Up Leina in Your Chat → -
Check the Result
Review the result and any changes in the connected app.
Supported Dandelion API Capabilities
Dandelion text analysis connects entity mentions to references and estimates language, sentiment, and semantic similarity. These outputs support document organization and text comparison.
- Analyze Plain Text
-
Extract linked entities and text positions, detect languages, analyze sentiment, and compare two texts’ semantic similarity.
- Extract Entities
- Detect Language
View All 4 Actions
- Extract Entities
- Analyze Sentiment
- Detect Language
- Compare Text Similarity
Frequently Asked Questions
Can I submit a document link?
These operations take plain text. Provide the relevant text rather than only a document URL.
Can an entity index point back to the original text?
Yes. Entity extraction provides positions that can be kept alongside the entity references.
Does a similarity score prove that two texts are identical?
No. It measures semantic similarity; ask AI to compare the wording and specific claims as well.
Your AI employee for Dandelion API
Tell Leina what you need in your team chat. Include the source, destination, and expected result.