Entity Extraction with Rosette

100%

Entity Extraction with Rosette

✓ Published0🌍 Public

Last edited Aug 18, 2017

Created on Aug 18, 2017

This example shows entity extraction on archival finding-aid text, displaying the JSON response from the Rosette API for each `<ead:item>` element. The XQuery script fetches a remote XML document from GitHub, serializes each item’s normalized text into a JSON payload, and posts it to Rosette’s `/entities` endpoint. It uses the EXPath `http:send-request` function with an `<http:header>` for the API key. The query filters responses to those with entity results, but the code does not visualize them, serving instead as a data-gathering pipeline.

AI-generated description

Similar vizzes

Loading thumbnail…

XQuery to clean up data from an Excel-generated webpage & generate an EAD document

This example shows how XQuery 3.1 can clean data scraped from an Excel-generated HTML table and use it to generate an EAD (Encoded Archival Description) finding aid. The `fetch:text()` and `html:parse` functions extract rows from a legacy webpage, normalize the text content, and repackage the fields into JSON maps. A second function transforms those entries into EAD `<c02>` components, inserting metadata like titles, abstracts, and thumbnail image links from the Lantern Slide collection at Vanderbilt. Finally, the script rebuilds a complete EAD document, injecting the new `<c02>` elements into the existing archival description hierarchy, preserving the original structure while enriching it with parsed data.

CCliffordAnderson

63% match

Loading thumbnail…

Extracts entities with geographic attributes using the Alchemy api

This example shows geographic entities extracted from a web-accessible text, filtered to include only those with associated geo-coordinates. The XQuery script sends a POST request to the Alchemy API’s URLGetRankedNamedEntities endpoint, supplying an API key and a target URL, then parses the returned XML to select entities that contain a `<geo>` element. The visualization relies on the Alchemy API for entity extraction and geographic tagging, while the code itself handles the HTTP request and response processing without additional rendering libraries. The data source is a plain-text file hosted on GitHub, and the output is a list of entities with location data.

CCliffordAnderson

62% match

Loading thumbnail…

Get finding aid

This example demonstrates how to fetch and transform an XML finding aid from a GitHub repository, isolating a specific portion of its hierarchical structure. The code defines a function that rebuilds the EAD document by extracting the header, archdesc attributes, and selected child elements while discarding others. It uses the XQuery language with a custom function `local:build-ead` to restructure the XML. The data source is a remote finding aid document loaded via `fn:doc` from a raw GitHub URL. The transformation explicitly copies the `c01` element and its preceding siblings, omitting the `c02` children, thereby illustrating selective XML reshaping for archival metadata.

CCliffordAnderson

60% match

Loading thumbnail…

Serializes disambiguated entities to CSV

This example extracts named entities from a web-accessible text and serializes only those with geographic coordinates into a CSV file. It uses the Alchemy API’s `URLGetRankedNamedEntities` endpoint via an HTTP POST request, then filters the returned XML for entities containing a `geo` field. The code constructs a `<csv>` element with each record’s text, longitude, and latitude, and outputs the result using BaseX’s `csv:serialize` function. The data source is a raw text file from GitHub, and the output is a clean CSV suitable for mapping.

CCliffordAnderson

59% match

Loading thumbnail…

Reads a page from a given book at the Internet Archive and extracts disambiguated geographic entities as GeoJSON using the Alchemy API

This example shows how to transform a digitized book page from the Internet Archive into a map of geographic locations mentioned in the text. It extracts named entities from page 15 of a scanned text and filters for those with geographic coordinates, then renders them as point features on a map. The code uses the Alchemy API’s `TextGetRankedNamedEntities` endpoint to perform entity recognition and disambiguation, while XQuery and BaseX’s `json:serialize` handle the XML-to-GeoJSON conversion. The output is a FeatureCollection with point geometries, built from the disambiguated `geo` coordinates returned by the API, with each feature’s properties containing the entity’s text.

CCliffordAnderson

58% match

Loading thumbnail…

Get Zotero JSON from API

This example demonstrates how to retrieve and parse Zotero library items as structured XML from the Zotero API. It uses the XQuery `fetch:text` function to download JSON data from the API endpoint, then parses that JSON with `json:parse`. The code extracts all `note` elements from the parsed data and converts each into an XML fragment using `fn:parse-xml-fragment`. This approach shows a direct pipeline from API response to manipulable XML nodes, useful for further processing or transformation in an XQuery environment.

CCliffordAnderson

57% match

Loading thumbnail…

Simple example of retrieving a record in EAD (or Dublin Core) from the Archives Hub using SRU

This example demonstrates how to retrieve a single metadata record from the Archives Hub using the SRU (Search/Retrieval via URL) protocol, converting it into a specified format such as EAD or Dublin Core. It shows a live HTTP query to the Archives Hub API, where the user can adjust the search keyword and the desired metadata schema (e.g., "ead", "srw_dc", or "oai_dc"). The code uses XQuery 3.0 with the `http:send-request` function to fetch the XML response, then extracts the record data via the SRU namespace. The result is the raw metadata record, ready for display or further processing.

CCliffordAnderson

57% match

Loading thumbnail…

VRA website example

This example demonstrates a focused query of a VRA XML metadata record, extracting and displaying only the `locationSet` element from the source document. The code uses the eXist-db HTTP Client module (`http:send-request`) to fetch the remote XML file from GitHub, then applies an XQuery path expression to isolate the location data. The visualization shows how structured art metadata can be programmatically filtered to reveal just geographic or provenance-related information. By leveraging the VRA Core 4 namespace declaration, the script ensures precise element selection, making it a practical tool for targeted data extraction in digital humanities projects.

CCliffordAnderson

57% match