Skip to content

Duplicate Detection

GET /api/levenshtein

Finds the closest stored publication title using PHP's similar_text comparison. Despite the endpoint name, the current implementation does not use Levenshtein distance.

Parameters

Parameter Type Description
apikey string API key, if required. Alternatively use the X-API-Key header.
title string Required publication title to compare.
pubmed string Optional exact PubMed lookup; its result is currently overwritten by the title comparison.
doi string Optional exact DOI lookup; its result is currently overwritten by the title comparison.

Example Request

1
GET /api/levenshtein?apikey=YOUR_API_KEY&title=Example%20publication

Example Response

1
2
3
4
5
{
  "similarity": 100,
  "id": "665ef1234567890abcdef1234",
  "title": "example publication"
}

Returned Fields

Field Description
similarity Percentage rounded to one decimal place. Values below 50 are reported as 0.
id ID of the selected publication, or an empty string if none was selected.
title Stored title after lowercasing and removing HTML tags, or an empty string.

Response Behavior

The response is a direct JSON object without status, count, or data. Authentication failures still use the standard permission error envelope. limit and offset are ignored.

The comparison selects the title with the largest number of matching characters, then reports that match's percentage. A low score can therefore still return an id and title. Only activities with type=publication are considered.