Leo, Andrea,
As one of those heavily involved in the software reuse movement of the 80’s and 90’s that Leo mentions, I have to chime in here (for a while a had a car vanity plate that said “REUSEIT”). While much effort was focused on making it easier to find/discover “reusable” components (and today, “services” – i.e., reusable functionality not burdened by execution environment dependencies), much less attention was paid to the issue of characterizing the underlying environmental and operating environment scope assumptions implicit in all software. Software language and hardware/OS platform dependencies where obvious ones, but institutional environment (name spaces, legal/regulatory frameworks, currencies, user ids, security/permissions, etc.) and operational domain dependencies (like implicit ontologies in data models and object models) were more less obvious obstacles. We focused on being explicit about the domain dependencies and degrees of variability in the domain that a particular component supported, and used a semantic net-based library to characterize components for reuse. That helped in single-enterprise situations (like, say, DoD), but left the institutional context assumptions as obstacles. Another aspect of institutional environment dependencies were issues of liability and ownership associated with software components, as well as long term life-cycle evolution of components across institutional boundaries (a software component vendor didn’t necessarily share the same vision of the trajectory that a software component should take with all the existing customers of that component).
This experience played a large role in our development of the SCOPE model in NCOIC (www.ncoic.org). The basic issue when one talks “reuse” is being explicit about the range of conditions/contexts and purposes for which the entity in question can be “safely” reused. Then there is the issue of what the potential addressable market might be that satisfies the range of conditions and purposes under which the entity can be reused. Finally, the issue of what portion of that addressable market the entity actually “penetrates” or captures, and what the business model is for re-capturing investment in reusability, determines whether the investment in a reusable component is worthwhile. Note that I am not excluding the notion of “social” investment such as open source projects here. However, most people will not expend a lot of resources on something if they see no apparent pay-off, even if that pay-off is just satisfaction in seeing other people use their work.
Another aspect of reuse that makes it difficult to achieve is that even if you try to be explicit about your scope assumptions, the potential “reusers” almost certainly have implicit scope assumptions you didn’t consider (as Rich Cooper will no doubt agree). These get “discovered”, usually in a somewhat unpleasant way, when someone attempts to reuse something that violates one or more of these scope assumptions (like the first Ariane V launch). Fortunately, it is usually less expensive to fix those “discovered” problems than to write the code from scratch – but you end up with a new variant of the code, which obviates some of the expected benefit of reusable components, unless the originator agrees/decides to incorporate your changes and they don’t break someone else’s reuse of the same component.
I happen to believe that all these same issues apply to the use and “reuse” of ontologies, which is why I stress being explicit about the scope assumptions of ontologies. The NCOIC SCOPE model wasn’t developed explicitly for characterizing ontologies, but the “O” in the SCOPE acronym could just as easily have stood for Ontologies (or business Objects), as opposed to “Operations”, which it does stand for.
Hans Polzer
From: ontology-summit-bounces@xxxxxxxxxxxxxxxx [mailto:ontology-summit-bounces@xxxxxxxxxxxxxxxx] On Behalf Of Obrst, Leo J.
Sent: Monday, January 27, 2014 4:43 PM
To: Ontology Summit 2014 discussion
Subject: Re: [ontology-summit] [Reusable Content] Characterizing or measuring reuse
Hi, Andrea,
María Poveda-Villalón presented during Ontology Summit 2013, and talked about OOPS!: at least for OWL ontology validation; see: http://ontolog.cim3.net/cgi-bin/wiki.pl?ConferenceCall_2013_01_31.
One issue we may not have raised yet (and I don’t want to go too far off on a tangent) is that reuse is not just an issue for ontologies (and other kinds of models, for that matter), but for software. Software reuse is not really that emphasized anymore (say, as opposed to a slight trend in the 1980s-1990s which promoted reuse). Consider object-oriented programming: one of the premises of OO was that, by having program constructs “closer” to real-world objects and properties (and data abstraction, polymorphism), greater design clarity and reuse would be promoted.
However, in my experience, programming code is nearly always developed from scratch, perhaps reusing via (string-based) cut-and-paste. There are software libraries (ostensibly certified) that some programmers do reuse and extend, and of course plugins/extensions based on open code (Firefox, Protégé, etc.) Service-Oriented Architecture has been a movement during the last 10+ years that tries to focus on individual reusable services, and it has had some success.
The rise of maturity models in information technology and model-driven architecture tries to address software reuse, to some extent. But the “not-invented-here” syndrome afflicts software, and perhaps also ontology development.
Thanks,
Leo
Other important questions in the "reusable content" arena are how to ascertain and improve the amount of reuse.
It "seems" that reuse is low, but there are many sites offering reusable content and therefore many opportunities for reuse. For example, in the Ontology Design Pattern (ODP) space, there are:
- W3C'S Ontology Engineering and Patterns Task Force (OEP) [1]
- Ontology Design Patterns org wiki [2]
In addition, there are foundational ontologies available, as discussed in the Upper Ontology Summit (2006) [4], as well as domain ontologies like FIBO.
So, does the wealth of information contradict the perception?
Or, is content present but it is just very difficult to use/re-use?
Perhaps we need to refine our engineering approaches and abilities to better find and evaluate reusable content? This is discussed in a paper by María Poveda-Villalón, Mari Carmen Suárez-Figueroa and Asunción Gómez-Pérez [5] that I found quite interesting.
I personally would love to see a review and recommendation system put in place for ontologies, patterns, linked data models, etc. Is this something that we could achieve?