Qualifying with AsTMa=

Topic Maps can capture facts via associations. Whenever one or more topics play roles in a statement, we define an appropriate type characterizing the relationship and insert the topics into their respective roles.

As a simple example, let us look at "France Telecom controls at least 30 percent of the swedish phone market". In AsTMa=, we would code this as

    # statement 1
    (holds-share-in-market)
    holder: france-telecom
    share: percent-30-gt        # 30 % or greater than
    market: swedish-phone-market

In many applications it is also necessary to make qualifications about these statements. We might be unsure about their truth value, would like to attribute it to a particular source or any other reason why the statement should be weighted in a new context.

Elaborating on our example, we might want to add that the above statement is supposed not to hold now and for all times but only for a particular time frame: "In the second half of 2002, France Telecom will control at least 30 percent of the swedish phone market." In a first attempt, we sketch

   # statement 2
   (prognosis-timeframe)
   timeframe: time-y-H202   # this should encode the 2nd half of 2002
   prognosis: ????

But before we can insert statement 1 for the ?'s, we have to create a topic from our earlier association. This process of reifying an association is encoded as follows:

    # statement 1a
    (holds-share-in-market) is-reified-as statement-1a
    holder: france-telecom
    share: percent-30-gt
    market: swedish-phone-market

AsTMa= will:

• create this association like in statement 1,
• create a topic with id statement-1a, and
• create a subject indicator with a reference pointing to the association.

Once statement-1a has been created, it can be used straightforwardly in

   # statement 2
   (prognosis-timeframe)
   timeframe: time-y-H202   # this should encode the 2nd half of 2002
   prognosis: statement-1a

Along these lines we can then add more and more qualifications, such in "According to a survey, it is possible, that in the second half of 2002, France Telecom will control at least 30 percent of the swedish phone market."

   # statement 2a
   (prognosis-timeframe) is-reified-as statement-2a
   timeframe: time-y-H202   # this should encode the 2nd half of 2002
   prognosis: statement-1a
   probability: possible    # however this is quantified

   (is-source-of-information)
   source: some-survey
   information: statement-2a