The visualizations of the Dive into Intangible Cultural Heritage project are built upon the foundation of an extensive and open dataset. This dataset is updated once a year after the session of the Intergovernmental Committee for the Safeguarding of the Intangible Cultural Heritage (ICH). It is released by UNESCO according to the principles and core values of open science.


The dataset is organized as a semantic graph, where nodes (also known as vertices) are connected by edges.

The most important nodes in this graph are the elements inscribed in the three ICH lists (Representative List, Urgent Safeguarding List and the Register of Good Safeguarding Practices). Other types of nodes include projects, NGOs, countries, regions, World Heritage Convention elements, case studies, scientific publications and Sustainable Development Goals. Another node type acts as a metaphorical glue between all these vertices, giving structure and meaning to the whole graph: this is the concept type. These concept terms are taken from various thesauri, but mostly from the official UNESCO thesaurus and a specialized vocabulary curated by the ICH Secretariat. For example, both the Royal ballet of Cambodia and the Capoeira Circle elements are connected to the Dance concept.

The complete graph structure is detailed below.


The full dataset is available as a JSON graph. For convenience, subsets of the data are also provided as self-explanatory CSV tables.

Full datasetDownloadDownloadDownloadJSON
Sub dataset - ConstellationDownloadDownloadDownloadCSV
Sub dataset - Sustainable Development GoalsDownloadDownloadDownloadCSV
Sub dataset - SDG #1: No povertyDownloadDownloadDownloadCSV
Sub dataset - SDG #2: Zero hungerDownloadDownloadDownloadCSV
Sub dataset - SDG #3: Good health and well-beingDownloadDownloadDownloadCSV
Sub dataset - SDG #4: Quality educationDownloadDownloadDownloadCSV
Sub dataset - SDG #5: Gender equalityDownloadDownloadDownloadCSV
Sub dataset - SDG #6: Clean water and sanitationDownloadDownloadDownloadCSV
Sub dataset - SDG #7: Affordable and clean energyDownloadDownloadDownloadCSV
Sub dataset - SDG #8: Decent work and economic growthDownloadDownloadDownloadCSV
Sub dataset - SDG #9: Industry, innovation and infrastructureDownloadDownloadDownloadCSV
Sub dataset - SDG #10: Reduced inequalitiesDownloadDownloadDownloadCSV
Sub dataset - SDG #11: Sustainable cities and communitiesDownloadDownloadDownloadCSV
Sub dataset - SDG #12: Responsible consumption and productionDownloadDownloadDownloadCSV
Sub dataset - SDG #13: Climate actionDownloadDownloadDownloadCSV
Sub dataset - SDG #14: Life below waterDownloadDownloadDownloadCSV
Sub dataset - SDG #15: Life on landDownloadDownloadDownloadCSV
Sub dataset - SDG #16: Peace, justice and strong institutionsDownloadDownloadDownloadCSV
Sub dataset - SDG #17: Partnerships for the goalsDownloadDownloadDownloadCSV


The graph is represented as a JSON object with three main sections. The meta key holds generic information such as the language and the timestamp of the dataset. The nodes object contains detailed information about each vertex. At a minimum, a node is defined by a type and a label. Additional keys may be available according to the type. For example, the concept type has an additional group field that maps to an identified thesaurus. Most other types define a meta object that further describe the item. The full list of properties can be found below. Finally, the edges array link all the vertices together in the form of RDF tuples. However, there is an additional property to the traditional triple: the weight key, which gives more or less importance to a connection.


"meta": { # graph metadata
"language": <string>, # en|fr|es|ar
"generated": <string> # YYYY-MM-DD HH:MM:SS
"nodes": { # nodes main object
<id>: { # node ID
"type": <string>, # element|project|ngo|country|region|concept|whc|casestudy|publication|sdg
"group": <string>, # unesco|main|biome|nature|threat|domain (concept)
"label": <string>, # node title
"meta": { # node metadata, type-dependent, displayed on click or tap (popover)
"icon": [ # node main image URLs (element, casestudy)
"small": <string>, # small version
"large": <string>, # large version
"description": <string>, # description (element, project, ngo, whc, casestudy, publication)
"list": <string>, # RL|USL|GSP (element),
"year": <integer>, # inscription year (element),
"multinational": <boolean>, # whether it is linked to more than one country (element)
"link": <string>, # external link (element, project, ngo, whc, casestudy, publication)
"language": <string>, # language of bibliography entry (publication)
"paper": <string>, # Full paper URL (publication)
"keywords": [ # Keywords (publication)
"images": [ # list of associated images (element, casestudy)
"url": <string>, # image URL
"copyright": <string>, # copyright
"title": <string> # title
... # up to 10 images, ordered by relevance
"video": { # associated Youtube URL (element)
"url": <string>, # video URL
"copyright": <string>, # copyright
"title": <string> # title
"sustainability": <string> # Information about sustainability (element)
"considerations": <string> # Considerations for inclusion (sdg)
"edges": [ # edges main array
"subject": <id>, # source node ID
"predicate": <string> , # broader|narrower|related|exactMatch|closeMatch|primeExampleOf
"object": <id>, # target node ID
"weight": <integer> # 3 (primary concepts) | 2 (secondary concepts) | 1 (all other nodes)