[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]
Subject: Re: thoughts on scope
So typically I would expect someone to provide search criteria (say, property/value targets) against a corpus and get back those instances that best match the criteria. Then the question is whether there is value in filtering for instances that are described by my properties of interest (i.e. have a compatible scope) before ordering against my target values. Let's look at another case. I identify 10 properties and corresponding target values and there are two instances: one that exactly matches 5 target values but doesn't include the other 5 in its description; a second instance is described by all 10 properties but only has a mediocre fit to the target values. Which instance do I want as the best fit? If the latter, then having scope is useful. If the former, I'd argue that scope doesn't provide me much. In reality, I'd argue it depends on the relative importance to my task of each of the properties and it is really some combination of what information is provided and what the information tells me that determines best fit. For this, I'd say scope as you define it provides little value. Ken On Sep 26, 2007, at 8:37 PM, Francis McCabe wrote:
----------------------------------------------------------------------------- Ken Laskey MITRE Corporation, M/S H305 phone: 703-983-7934 7151 Colshire Drive fax: 703-983-1379 McLean VA 22102-7508 |
[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]