Initial question 4: What should NSCS codes look like?

Archived Page.

This page related to stage 1 of the project, which concluded in September 2014.

While the new subject coding scheme’s structure could accommodate a range of codes or identifiers, the shape of the most prominent codes can still make a big difference to useability. Read through the issue and the solutions, and let us know what you think the solution should be, and why.

The issue:

Since disciplines can change their names, or be known under different names by different groups, it can desirable to have codes or identifiers for terms that are entirely independent of their names or labels. Furthermore, if terms are to be shared between different hierarchies, identifiers that encode hierarchies can become problematic.

Solution 4a. hierarchy in the codes

Assuming a hierarchy, 4a shows the broader terms of a low level term in the code that identifies it.

JACS3 codes

JACS3 codes

Pros:

  • allows for a degree of label – identifier differentiation
  • supports recognising the context or hierarchy of a term

Cons:

  • tricky to balance a wide value space with recognisable codes
  • difficult and misleading when sharing terms between hierarchies
  • can lead to shifting value problems over time

Solution 4b. label-derived codes

In 4b, a shortened version of a widely used label is used. This solution works with either flat or hierarchical coding schemes.

Pros:

  • very easy to use
  • avoids coding mistakes
  • supports term sharing

Cons:

  • can lead to problems after shifts in meaning or labeling
  • potential increase in keying errors because of length of term or misremembering abbreviation
  • does not easily support context recognition in the code
  • can be confusing when used with local labels
  • may be difficult to use abbreviations for very long labels, for example Eastern, Asiatic, African, American & Australasian languages, literature & related subjects

Solution 4c. random codes

In 5c, terms get assigned a random, short code that just identifies the term, but carries no meaning itself. This works as well with hierarchies as with flat lists.

Codes with no intrinsic meaning

Codes with no intrinsic meaning

Pros:

  • allows for shifts in labelling over time
  • supports term sharing
  • limitless value space

Cons:

  • can be hard to use in practice
  • difficult to spot errors

Solution 4d. Hierarchy in the codes, with space (variant of 4a)

This solution has a hierarchy of codes (i.e. structure within the notation), but uses well-spaced codes, so that new codes can be added, or codes can be split easily. It uses a spacer for readability and to indicate level, in this case a dot. All Level 1 codes would have nnn format, Level 2 nnn.nnn, Level 3 nnn.nnn.nnn.

code2

Or the notation could be shortened, using 2 letters instead of digits.

code3

The idea here enables easy addition of new terms without disrupting existing notation:

code4

and also splitting of terms for greater differentiation, if needed, again without disrupting existing notation. In this example, the two new terms could even be added to the hierarchy as a new 4th level, extending the notation to 8 letters.

code1

Pros:

  • allows for a degree of label – identifier differentiation
  • supports recognising the context or hierarchy of a term
  • enables development of the schema over time without wholesale re-structuring
  • allows for addition of, or linking to, more detailed levels than the original framework
  • allows for shifts in labelling over time
  • limitless value space

Cons:

  • codes are not memorable
  • danger in some cases that users might expect meaning from the letters (for example as abbreviations of the terms)
  • length of codes not consistent, as they depend on level (though this could be mitigated by using ZZ as a blank, for example EK.RK.ZZ as a level 2 code).

 

View more initial design questions on the Design page

Leave a Comment

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s