I recently posted here (1) and here (2, better) about the pesky lack of an ontology for physical activity and its effect on fitness. Ontology is a broad and somewhat ill-defined term, but here I mean unambiguous language for describing physical exercise as well as its effects on specific domains within fitness and wellness. Much like we have unambiguous language for describing drug dose and response. My belief is this prevents us from having more interesting and useful activity trackers.
Has anyone any experience with this subject? Has anyone ever come across such an ontology?
I have experience with ontology both in the philosophical sense and in the mathematical sense of the word. I also have experience with controlled vocabularies (from bioinformatics).
I am not aware of any ontology for biometrics, but I agree that one would be useful, and could seriously improve integration between devices and bring down some of the closed corporate walls. Such an ontology would have to account for the different sensor sources and for “black box” inferences from many of the propitiatory technologies in use as so many companies don’t like to disclose their algos.
I really appreciate this discussion and agree with Joshua that the black box algorithms make ontologies challenging, as step does not equal step’. Even different versions of the same service may disagree after firmware upgrades…
Thanks for the reference. Just checked out your project.
You connect target muscle groups to specific exercises. This makes sense if your goal is to make certain muscles bigger or stronger.
However, what if your goal is to improve “heart health”, or “insulin sensitivity”, or “cognitive performance”? While we know exercise improves heart health, it is tough to describe exercise in a way that is consistent with optimizing heart health. Rather than focusing on just a muscle group, an exercise would probably have to be described in terms of duration, heart rate, maybe blood oxygenation, etc. The difference is similar to how a body-builder is interested in the protein and calories in a certain food item, while a long-distance runner might be interested in the same food’s glycemic load.
Further, I think improving things like “heart health” or “cognitive performance” are more interesting than muscle building/strengthening because they fit more into an overall goal of wellness and disease prevention.
I personally am working on the cognitive performance outcome. I’d be interested in collaborating if you wanted to focus on that specific outcome…
I have read many studies linking strength training with improved diabetes and to reduce cardiovascular risk.
My idea is build a system capable of generating training routines as user goal and fitness. So that, if you are a beginner and just want to improve their health, the system generates a weight training oriented to it. However, if user wants to gain strength, the system will indicate what exercises should be performed, number of repetitions, rest periods, etc.
I started with the creation of ontology separating exercises per muscle group, user level (beginner, intermediate or advanced), type of exercise, necessary equipment … and later will create ontologies with user data and routines that can be generated depending on user needs.
I think that ontology could be used for your projects also
In relation of diet, I have studied nutrition also but in my system… I don’t want to extend to this subject, because I want to test system with real people and If I extend a lot, I couldn’t finish the system at time… But nutrition section could be a future work on this system
I highly suggest you read Dr. McGruffs book Body By Science. The dosage of reps and etc. would not be eliciting the proper biochemical and epigenetic stimulus to really improve their fitness/health. Yes fitness is normally partnered in thought with health. But its diminishing returns. Super fitness does not equal super health. Findings in evolutionary biology it can be seen that there is a universal stimulus that once triggered initiates the muscular growth processes. It takes much much less time than mainstream thought.
