The WordCorr Project


Home ] [ Background ] [ Technical ] [ SourceForge Project Page ] [ Download ]


IPA by keyboard ] [ IPA by zone ] [ Use Cases ] [ Test Data 1 ]

Functional Test Data 1

To view the test data you need the Arial Unicode MS font installed on your computer. Download it from Microsoft.

Incoming data are from the keyboard, or from an importation process along with other forms that may need to be typed in to supplement the importation. I am using informal XML notation to highlight the kinds of data and their interrelationships, not to lock them into a particular form. The XML codes indicate what kinds of screens are necessary, and hint at their behavior.

Once the data are inside WordCorr, they are displayed for annotation under a particular view: for each entry, each line gets a tag to show which group of comparable forms it is to be tabulated with for that view, and all members of a group get their glyphs aligned by insertion of Indel (insertion or deletion, don't know which yet) symbol glyphs to show that something has come in or dropped out, and by insertion of Ignore symbol glyphs to show that the glyph in that position is not to be compared because it is already recognized as extraneous on linguistic grounds. All aligned data in a single group must have the same number of glyphs. There is a four-component vector that defines metathesis if any. There may be a remark of any size with any datum or aligned datum.

The raw data are displayed according to the XML pattern shown, and alignments and annotations edited into the display by the linguist. For a data keeper or team leader, the same display allows editing of the raw data.

As data groups are tabulated as correspondence sets, the correspondence set for each glyph position has the content shown. Each correspondence set is placed immediately in Results for the view, using for initial placement information supplied at that time by the linguist about the protosegment and environment.

Results, which can be refined further, holds the correspondence sets as evidence in a structure shaped according to the linguist's hypotheses. A later release will give a more elaborate presentation suitable as an appendix to a monograph or book.

For problems or questions regarding this web contact khamasak@users.sourceforge.net.
Last updated: Jan 01, 1970


Sponsors:

SourceForge.net Logo Data House, Inc. Logo University of Hawaii Logo NSF Logo SIL Logo