Re: [UCR] Managing Inter-Organizational Business Policies & Practices: Edited Version.

What I would like to do now is focus on the well-articulated 
requirements you mentioned at the f2f that this use case describes. My 
main worry about the use case is that it coudl be interpreted as 
dictacting requirements for RIF that are out of scope. So most of my 
reservations woudl be alleviated by focusing on the specific 
requirements (this is something we should now be doing for all the use 
cases) and then, as with the other use cases, re-evaluating how well the 
use case itself captures and reflects these requirements.

-Chris

John Hall wrote:
> Hello Chris,
>
> My responses in-line, below,
>
> Regards,
>
> John
>
> Chris Welty wrote:
>>
>>
>> The new draft of this use case is much better, but there are still 
>> parts in it that border on what I initially objected to. Again, I 
>> think RIF is clearly about machine-processable rules. Of course if 
>> people want to use it for other things, great, I won't stop them, but 
>> these other things (like writing regulations or laws) shoudl not 
>> influence the design of RIF.
>>
>> Most of the current version is OK, I think, except:
>>
>> "EU-Rent UK finds some problems in applying the rules. One is that 
>> sometimes it has to give free upgrades to customers. It wants to have 
>> one of the rules for insurance tax changed.
>>
>> For an individual rental, the tax on aggregated insurance is 1.5% of 
>> the simple cost of the rental actually paid by the customer (not the 
>> price of rental of the upgrade provided)
>> It provides this to AutoLaw, which will negotiate it with the 
>> regulators and disseminate the outcome to EU-Rent and its other 
>> customers. "
>
> The issue here is ambiguity. The rule says, “Insurance tax must be 
> charged on rental at 1.5% of rental-simple-cost”. Rental-simple-cost 
> in EU-Rent agrees with the regulator's definition: "price charged for 
> use of the car plus compulsory insurance, excluding extra charges such 
> as additional, drivers, optional insurances, additional equipment." 
> But in EU-Rent it has two specializations.
>
>     * Rental-actual-simple-cost: "rental-simple-cost that is actually
>       paid"
>     * Rental-tariff-simple-cost: "rental-simple-cost that is quoted in
>       the tariff for the car group used in the rental" (which is
>       higher than the price paid if a free upgrade is provided)
>
> EU-Rent UK wants the rule disambiguated. But it provides its preferred 
> version: “Insurance tax must be charged on rental at 1.5% of 
> rental-actual-simple-cost”, so that tax would be due on the price paid.
>
> Of course, the regulator might decide that it should be: “Insurance 
> tax must be charged on rental at 1.5% of rental-tariff-simple-cost”, 
> so that if a free upgrade is given, tax would be due on the benefit 
> provided.
>
> This can’t be the only kind of circumstance where the tool behind a 
> RIF client needs to request clarification of a received rule that is, 
> for its local vocabulary, ambiguous. One reasonable way of resolving 
> ambiguity is to propose a non-ambiguous alternative. Are you 
> suggesting that the RIF should not support this capability?
>
>>
>> This is precisely what I would live to avoid. RIF is not a format for 
>> exchanging legal language between people so that they can negotiate.
>>
>> Also:
>>
>> "It also has some existing insurance policies in place. They provide 
>> third-party insurance as an explicit item, and EU-Rent UK cannot get 
>> refunds on early termination. It asks corporate HQ for rules:
>>
>> Cost of third-party insurance will be built into the basic cost of 
>> each rental, unless there is an alternative insurance already in place"
>>
>>
>> I do not see how this rule could be implemented in RIF, in particular 
>> "will be built in". What is that supposed to mean? This is not a 
>> machine processable rule.
> The issue here is about rules for building and using rental tariffs. 
> EU-Rent HQ wants the cost of 3rd-party insurance to be included in 
> tariff rates (the rates quoted for car groups and rental periods, e.g. 
> 'one-day rate for mid-size', 'weekly rate for full-size). EU-Rent UK 
> wants its current 3rd-party insurance, with distinct premiums for 
> 3rd-party insurance, to run its course before applying the new rules.
>
> Rules that would be machine processable might (given a more formal 
> syntax and a suitable vocabulary) look something like this:
>
>     * If 3-P insurance does not exist then aggregate-insurance must
>       exist and current-date must be or be later than
>       effective-start-date of aggregate-insurance and current-date
>       must be or be earlier than effective-end-date of aggregate-insurance
>     * If current date is later than (effective-end-date of
>       3P-insurance - 15 working-days) and current-date is or is
>       earlier than effective-end-date of 3P-insurance then
>       aggregate-insurance must exist and effective-start-date of
>       aggregate-insurance must be effective-end-date of 3P-insurance
>     * If current-date is (effective-start-date of aggregate-insurance
>       - 10 working days) then process “tariff update” must be
>       activated with effective-start-date of tariff =
>       effective-start-date of aggregate-insurance.
>
> Rules relevant to process “tariff-update”:
>
>     * If effective-start-date of tariff is earlier than
>       effective-end-date of 3P-insurance then tariff-item (car-group,
>       rental-period) <= base-rental-cost (car-group, rental-period)
>     * If effective-start-date of tariff is or is later than
>       effective-start-date of aggregate-insurance then tariff-item
>       (car-group, rental-period) <= (base-rental-cost (car-group,
>       rental-period) * (1 + aggregate-insurance-percentage (car-group,
>       rental-period))
>
> Rules relevant to process “rental quote”:
>
>     * Current-tariff is tariff with latest effective-start-date that
>       is earlier than current-date.
>     * If effective-start-date of current-tariff is earlier than
>       effective-end-date of 3P-insurance then rental-simple-cost <=
>       tariff-item (car-group, rental-period) of current-tariff +
>       3P-insurance-premium (car-group, rental-period)
>     * If effective-start-date of current-tariff is or is earlier than
>       effective-start-date of aggregate-insurance then
>       rental-simple-cost <= tariff-item (car-group, rental-period) of
>       current-tariff
>
> I'm sorry if I expressed the rule so informally that it looked as if 
> it couldn't be automatically processed. I had hoped that people would 
> get the idea without the tedious detail. The more machine-friendly 
> version above probably requires RIF WG members to understand more 
> about the detail of EU-Rent than they would ever wish to know.
>
> Or are you suggesting that effective date/times for rules should not 
> be supported by the RIF?
>>
>> What do others think here?
>>
>> -Chris
>>
>>
>


-- 
Dr. Christopher A. Welty                    IBM Watson Research Center
+1.914.784.7055                             19 Skyline Dr.
cawelty@frontiernet.net                     Hawthorne, NY 10532
http://www.research.ibm.com/people/w/welty

Received on Friday, 17 March 2006 01:16:08 UTC