Labels as data II

This section offers a different view of using labels as data and considers the differences in meaning they can convey. The following four models represent a person's letter writing record. Notice how "Writes" can play a different role in each one. Remember that any time a label is used in an object, the branch leading up to that label is also included in the object.


Model 1        ?Person Writes

Model 2        ?Person Writes UNIQUE No
                                     Yes

Model 3        ?Person Writes UNIQUE Annually
                                     Monthly
                                     Weekly
                                     Daily
                                     Hourly

Model 4        ?Person Writes Fun_letters
                              Long_letters
                              Perfumed_letters
                              On_colored_paper
                              When_they_want_somthing
                              About_their_cats

The first model represents letter writing as a "unary" concept. If "Writes" is present it means the person writes letters. Absence is ambiguous and could mean the person doesn't write or there is no data.

The second model is "binary"; it represents an either/or situation. "Writes" will be present for both "No" and "Yes" objects. A query on "Writes" might be used to identify people for whom letter writing status is truly known.

The third model supports a "fuzzy" notion of letter writing. "Writes_me_letters" is used here to imply that a person writes to some extent, the frequency varying across some spectrum.

The final model is even fuzzier since "Writes" might imply that any or all of the categories that follow it apply.

It is possible to construct very complex trees using nothing but "cascades" of labels to represent more than one concept.

?Person Writes UNIQUE No
                      Yes Frequency UNIQUE Annually
                                           Monthly
                                           Weekly
                                           Daily
                                           Hourly
                          Qualities Fun_letters
                                    Long_letters
                                    Perfumed_letters
                                    On_colored_paper
                                    When_they_want_somthing
                                    About_their_cats

Entering this data:

Person Sarah
Annually
Long_letters
About_their_cats

Results in this display:

[Thanks to Otto Ritter for introducing the "unary", "binary" and "fuzzy" concepts to me]

Show me the whole t-shirt model again.

Back to Table of Contents