TagDoc:Ratings

The tag documentation pages of TagDoc contain a number of topics on which ratings are provided, both verbally in the text as well as in a five level rating - see TagDoc:Guidelines for details on the categories. These are typically qualitative assessments and not quantitative measurements. Here i explain the meaning of the five rating levels.

Geometry types

Where useful TagDoc provides these ratings separated by geometry type - that is:

• Points - that is tagged nodes in the OpenStreetMap data model connecting the tags with a single 2D coordinate pair or location
• Lines or Linestrings - that is tagged linear ways in the OpenStreetMap data model
• Polygons - that is closed ways representing simple polygons in the OpenStreetMap data model - or Multipolygon relations

The distinction if a closed way (first and last node identical) represents a linestring or a polygon is based on the presence of an area=* tag, or - in the absence of such - implicit defaults based on the tags of the feature. These defaults are subject to changing conventions - which are documented on TagDoc as well.

Data quality ratings

For the data quality categories (Consistency in use, Value accuracy, Geometric accuracy)

• excellent - can be used for high quality applications without significant constraints.
• good - can be used but high quality or otherwise sensitive applications might require preprocessing or supplementation with other sources of information to improve quality or eliminate errors.
• moderate - is of use for many data users but applications will need to deal with quality issues on a regular basis or have low quality demands.
• bad - data can still be of use for some applications but in most cases only supplementary to other sources of information.
• very bad - data is so low quality that it is not of practical use for any serious applications.

Verifiability rating

For the Verifiability ratings the levels have the following meaning

• excellent - verifying the tag or the geometry it applies to is typically without problems.
• good - verifying the tag or the geometry it applies to can require more thorough scrutiny or more elaborate investigation but is typically still reliably possible.
• moderate - verifying the tag or the geometry it applies to is practically not fully possible, it requires relying on external, not independently verifiable secondary sources of information or making a subjective assessment. Core elements of the information this tag and the corresponding geometry transport however are typically still verifiable.
• bad - the tag or the geometry it applies to is predominantly non-verifiable, key elements of the mapping represent a subjective assessment or some authoritive data that cannot be verified independently.
• very bad - the tag or the geometry it applies to is typically not verifiable, it represents a subjective assessment or some authoritive data that cannot be verified independently.

Completeness rating

For the Completeness ratings the levels have the following meaning

• excellent - mapping is without larger systematic gaps in the area in question.
• good - there are gaps but overall a very solid percentage of existing features is mapped.
• moderate - a significant percentage of existing features is mapped but there are large gaps.
• bad - There is a bit of mapping using this tag but it is patchy.
• very bad - there is only very sporadic mapping with this tag by a few individual mappers with no significant coverage overall in the area in question.