By Clive Gold, CTO Marketing, EMC Australia and New Zealand
Can we value data and should we bother?
At the SNIA event this week all the vendors demanded that we define the value of data! Well we all know that sounds easy, that sounds logical, but does it actually mean anything? I just started wondering how I would value my personal data. How would I put a value on the different types of data that I have.
Just to be transparent, I have about 1TB of photo’s and a little more video content because I keep everything! I do think my children are the most recorded individuals the world, but I suspect that this is fairly normal, today. So I do have a few, (read perhaps thousands), of photos that are not quite as in focus, not framed well and the lighting not that good. Why keep them? Well you try convince their mother that the slightly fuzzy image of her child is lousy and should be deleted. (Good luck!)
So firstly is the question of the value according to who? The individual, the department, the organisation, the country or the world! OK first reaction is how can data have relevance beyond the microcosm it was created in? Well have a look at http://www.wefeelfine.org/ were Jonathan has aggregated mentions of ‘feel’ on twitter and blogs, and you can now understand what the effects of weather, age, etc. are on the way people feel.
Second question is the value at what point in time? If the past is the best predictor of the future, then all data has value that grows over time. If we had good data over the last 100 years about the climate, perhaps there would be more certainty on cause, effect and consequences of these changes.
Third, the value as part of a ‘mashup’, or the value of this data when combined with other data. How do we estimate what the value of this particular data set when you can find new insights when you combine it with data.. that may not exist today. A great deal of the genetic research that is based on empirical evidence is the ability to do correlations. Looking for the genetic sequences that would indicate the propensity to particular illnesses.
I’m sure there are more scenarios but I’m sure these raise enough issues. So what do you think? I raised the issue at the SNIA event, IBM didn’t agree.. Said tape was the answer, but I didn’t think that was valid, and HP really didn’t commit. http://www.computerworld.com.au/article/395898/iiis_delete_delete_–_big_data_question