/research/researchworks/naco/default.htm

originally: http://www.oclc.org/research/researchworks/naco/default.htm

Learn more about the NACO Normalization Service

OCLC Research's NACO Normalization Service enables systems to convert names and other text strings to a format more conducive to machine comparison and sorting.

Try it out

Go to the OCLC Research NACO Normalization Service.

Enter the string of characters to be normalized. Typically this would be the name of an author or other entity associated with a work:

  • Alain-René Lesage
  • Lesage, Alain-René
  • U.S. Dept. of Agriculture

The service also can accept data in the MARCMaker format:

  • $aLytton, Edward Bulwer Lytton $cBaron $d1803-1873

Press <Submit Query>.

The demo will return the text string's normalized form:

  • alain rene lesage
  • lesage, alain rene
  • u s dept of agriculture
  • lytton, edward bulwer lytton\baron\1803 1873

Why we developed this

This service is used to prepare text strings for machine comparison and sorting, according to the NACO normalization rules. These rules were developed for comparing name data. An example of such a comparison is determining whether a particular name already exists in an authority file such as the OCLC Research LC Name Authority Service. However, any kind of text string can be normalized by this service. We have found it useful in working with title strings as well as names.

Why do we need rules for matching and sorting? Because machine comparison is very specific, and names, in particular, can be recorded in different ways. It is useful to have rules about how the names to be compared are constructed, e.g., what character set is used, how capital letters and spaces are used, and how to handle diacritics and punctuation. The NACO normalization rules are designed for this purpose.

The NACO rules have been implemented in various ways. We reconciled three different ways of automating them to develop the NACO Normalization Service.

This work has been used in the OCLC Research FRBR projects. OCLC Research's NACO Normalization Service is part of the MIT/HP DSpace product.

Resources for Testing NACO Implementations

The following resources are available (ZIP:142K/6 files) to assist you in testing your implementation of the NACO normalization algorithms:

  1. NACO.py - a sample implementation
  2. NACOnocommas.script - a test script that removes all commas from normalized fields
  3. NACOnocommas.check - expected responses for the NACOnocommas script
  4. NACO.script - a test script that leaves the first comma in the first subfield 'a'
  5. NACO.check - expected responses for the NACO script
  6. NACO.java - a sample implementation

Please note that each of the sample implementations will run themselves against the scripts and expected response files.

 

Terms and conditions

Use of this site is subject to OCLC's terms and conditions. By continuing past this point, you agree to abide by these terms.

Try the online demo
Send us your ideas

Additional information