Finno-Ugric Objects Dataset (ERM / MuIS)

Access & allowed use

This dataset is provided for:

• hackathon research and prototyping
• visualizations and models
• presentations and demo publications

Not allowed:

• commercial reuse
• re-publishing raw data outside this repository
• redistributing linked media in bulk outside their original source context
• treating uncertain metadata as verified fact without qualification

---

Citation requirement

Estonian National Museum / MuIS open data.
Finno-Ugric collection dataset, 2026.

---

Data structure

Main file

finno-ugric-objects-dataset.csv (or .xlsx)

This is the main table, containing one row per object metadata record.

---

Related tables (linked via object_id)

• finno-ugric-object-places.csv
• finno-ugric-object-materials.csv
• finno-ugric-object-collectors.csv
• finno_ugric_object_relations.csv

These tables represent one-to-many relationships and are designed for easier analysis.

---

Technical details

• Delimiter (CSV): , or ;
• Encoding: UTF-8

---

Row represents

In the main dataset:

→ one metadata record describing one museum object

Not:

• necessarily one perfectly normalized object record
• a complete or authoritative museum description
• a single interpretation of place, people, or date

---

Main dataset fields

Field	Meaning	Notes
object_id	unique object identifier	stable MuIS-based reference
object_url	link to object page	public source record
image_url	link to media page	availability may vary
museaali_number	museum inventory number	collection-specific identifier
title_clean	normalized object title	useful for grouping
material	raw material field	may contain multiple values
material_category	broader material grouping	manually assigned
description_type_label	type of description	e.g. comment, legend
description_text	free-text description	rich but uneven
places	place information	may contain multiple values
people_names	ethnonyms / associated groups	simplified field
collection_date_raw	original date expression	textual
collection_date_start	normalized start date	ISO format
collection_date_end	normalized end date	ISO format
collectors	collectors / fieldworkers	may contain multiple names
related_objects	linked objects	may be sparse
relation_type	type of relationship	heterogeneous

---

Related tables explained

1. Places

finno-ugric-object-places.csv

• one row = one object + one place
• extracted from the main places field

Used for:

• mapping
• geographic filtering
• place normalization

---

2. Materials

finno-ugric-object-materials.csv

• one row = one object + one material
• derived from material field

Includes:

• detailed material values (preserved)
• original variation retained

---

3. Collectors

finno-ugric-object-collectors.csv

• one row = one object + one collector

Used for:

• analyzing fieldwork
• mapping collector activity
• building networks

---

4. Object relations

finno_ugric_object_relations.csv

• one row = one object pair (object ↔ related object)

Used for:

• identifying object clusters
• reconstructing sets or collections
• network analysis

Note:

• only a small subset of objects contains relations (~5–6%)
• relations are likely intentional, not system-generated noise

---

About description fields

The description_text field is one of the richest parts of the dataset. It may include:

• acquisition narratives
• collection context
• fieldwork notes
• use descriptions
• provenance information

It is valuable for:

• text mining
• qualitative analysis

However, it is:

• uneven
• multilingual
• historically layered
• not normalized

---

About place and people fields

The places and people_names fields are simplified analytical extracts.

Important characteristics:

• multiple values may appear in a single field
• historical and modern place names may coexist
• most locations are in present-day Russia
• names often reflect Soviet-era transliteration
• values range from broad regions to specific villages

These fields are:

• useful for exploration
• not fully standardized authority data

---

About museaali_number

The museaali_number reflects archival grouping.

Example:

ERM B 193:1 → ERM B 193:13

This indicates:

• a shared collection or expedition
• multiple objects belonging to the same acquisition

Useful for:

• reconstructing collecting events
• grouping objects by expedition

---

About dates

The dataset includes both raw and normalized date fields.

• collection_date_raw → original textual form
• collection_date_start → normalized start date
• collection_date_end → normalized end date

Patterns include:

• full date ranges
• partial ranges
• single-year entries

Notes:

• dates refer to collection, not object creation
• year is almost always available
• precision varies

---

Limitations

• metadata quality is uneven
• image links may not always work
• multiple images per object may exist
• some fields are missing or partially standardized
• object titles vary in specificity
• material categories are analytical simplifications
• place names may require normalization
• duplicate object rows may exist
• description texts may contain historically sensitive terminology

---

General note

This dataset is designed for:

• exploration
• experimentation
• prototyping

It is not a fully normalized museum database, but a research-friendly extract.