Evidence, intelligence and big data

[Beardy]

A philosophical musing in the shower this morning prompts me to pose the question below.

First though, a definition of MY terms;
Evidence: legally collected information regarding a person’s activities that could be presented in a court of law
Intelligence: information gathered that might suggest that a person is of interest and in need of monitoring, probably with a goal of collecting evidence.
Big data: the analysis of information from various and multiple sources, about a person and/or their activities, usually involving machine learning and AI techniques to data mine links to other information.

We all know that Big Data is a thing, and not something we can ever put back in the box. If commercial organisations are using it openly you can bet your bottom dollar that other agencies are also using it.

Putting aside terrorist type activity, do you think that it’s reasonable for Law Enforcement Agencies (LEA) to intelligence gathered using Big Data techniques as a basis to investigate people identified thus, with a view to gathering evidence to convict them? Is there a list of crimes which you feel it would be ok for LEA to use such techniques and another where it would be ‘unfair’?

Ps; mods, hope this ok in the general Pub.

[Kim]

Dog fouling, obviously.

[Wowbagger]

Yesterday, on my way back from the park, as I walked through the church yard, I noticed that clean-looking water was leaking from a metal cover embedded in the path. When I got home, I phoned my sister, who is a member of the said church and does some gardening there, to let her know about the leak. I interrupted her half-way through a zoom meeting in which she and some other parishioners were indulging in Virtual Afternoon Tea. She alerted the vicar, who was there, to the leak, and he replied "Yes, I've just got the CCTV footage now. There's someone looking at it - I think it might be your brother!"

Clearly, Proverbs 15: 3 is relevant here. "The eyes of the Lord are everywhere, keeping watch on the wicked and the good."

[ian]

Doesn't that article just say they looked at CCTV footage? I think I'm only surprised to realise there's occasionally someone behind those cameras, most gaze into the distance like distracted cats.

The main problem with big data is the data. In my experience, the people who run the algorithms (all much of a muchness, tbh, the basics have been around for decades waiting for computing power to catch up) often don't fully understand the data, with is a more subjective thing, the biases, omission and problems inherent in that data and how it was collected. There's nothing inherently wrong with big data, and modern computing power means it's going to happen more.

Ironically, it's ultimately not a data science problem. Data is only as valuable as the interpretation.

[Kim]

Yesterday, on my way back from the park, as I walked through the church yard, I noticed that clean-looking water was leaking from a metal cover embedded in the path. When I got home, I phoned my sister, who is a member of the said church and does some gardening there, to let her know about the leak.

Good thinking.  Last time I reported a water leak, I made the mistake of using the water company's online web form, with handy fields for a description and map coordinates of the leak.  Apropos of ian's post above, Severn Trent amply demonstrated their inability to interpret data by dispatching Digby and Dougy to my home address a couple of days later.

[Polar Bear]

Indeed but the big problems with that are who is interpreting the data, how powerful are they and how will it be used.

A surveillance state where trials can be behind closed doors and you are not even allowed to see the evidence against yourself encourages the misuse of data gathered as information and used as evidence interpreted to mean whatever big brother wants it to mean.

Now I'm sure that never happens anywhere in the world ...

[Tim Hall]

Yesterday, on my way back from the park, as I walked through the church yard, I noticed that clean-looking water was leaking from a metal cover embedded in the path. When I got home, I phoned my sister, who is a member of the said church and does some gardening there, to let her know about the leak. I interrupted her half-way through a zoom meeting in which she and some other parishioners were indulging in Virtual Afternoon Tea. She alerted the vicar, who was there, to the leak, and he replied "Yes, I've just got the CCTV footage now. There's someone looking at it - I think it might be your brother!"

Clearly, Proverbs 15: 3 is relevant here. "The eyes of the Lord are everywhere, keeping watch on the wicked and the good."

Or The Gospel of St Arthur: "No wicked for the peaceful"

[fuzzy]

Evidence is something that helps to prove or disprove a theory- someone's guilt or innocence, that the earth is spherical/ flat etc. though often most thought of in relation to court proceedings.

Intelligence is basically information that has been corroborated (it's accuracy verified by other sources or means) and does not necessarily constitute evidence though could become so. For example, Fred tells you that it is Mrs Miggins cat that is shitting in your daffodils and Fred knows because he heard it from Dick whilst at the pub. A this is just rumour, this is information. If Fred explains that he knows this because he saw it happen, it would be intelligence once it is sanitised and subjected to a classification regime. Sanitisation is where the intelligence is recorded in such a way that the source cannot be identified. Classification is a 3 step process relating to the veracity of the source, the sources means of knowing the information and the usability of the intelligence (some intelligence is so sensitive that it cannot be acted upon on its own).

Using Big Data in investigations or operations? Why not. If you don't want the Feds using the info, don't allow it to be out there. If you allow someone access to your data, be vigilant about the boxes you tick/ leave ticked. If your data is released by an entity outside your granted permissions, litigate the bastards till they glow.

[ian]

There's a black box problem too – it's often not clear – indeed not possible to say – how an output is reached based on the input data. That makes human validation of outcomes difficult.

[Polar Bear]

The problem is that in many ways you have absolutely no control of your data out there.  Awareness came late once data had already been commercially and governmentally exploited and any protection apparently offered is wafer thin and full of holes.

If I want to be paid I need a bank account.  If I take cash even normally right minded folk will accuse me of being a tax-avoiding scumbag criminal.  Once I have a bank account I can limit my data exposure by using cash but the bank has a record of my withdrawals and increasingly retailers are not taking cash.  If I want to walk around outside there are literally thousands of cameras to track my movement.  If I was a driver there are literally thousands of cameras able to read car registrations on gantries, poles and in police (and probably other) vehicles.  And to pay any taxes due I need a bank account because HMRC don't accept wads of folding.

The law requires me to register with my local authority and be on the electoral roll and yet folk can buy this information.  My healthcare records are supposed to be private but we know that information is shared with private organisations.  The tax man knows how much my income is and who pays me it - that was self-evident when I filled out my tax return because that information was already there when I signed into my government gateway account.   Crapita know where I live and onenof my bank account's details as I have a TV licence because I am a law-abiding person and bought one when I bought a TV 15 months ago. 

All my utility providers both past and present have records of me including my name, address, some further level of contact if offered, and my bank details as do my Internet and mobile phone providers. 

Helping vulnerable people sort out benefits entitlements and / or to look for jobs, homes to rent, sort council tax payments, file papers to court, etc. inevitably has required them to have an online presence.  The surveillance state is well and truly here and is all pervasive and we can do nothing at all about this. 

I don't do loyalty cards but if I did then both the retailers and scheme operators such as Nectar will know what I buy, how much I spend and where I spend it.

If I send an enquiry email to a company about returning a faulty product under warranty then you can bet that Google will start pounding me with ads for similar products.

And I am only scratching the surface.  It is nigh on impossible to live outside eof big data and thus avoid the exploitation that comes with it by both the state and by $BigCos.

I am so looking forward to some level of normality again when I can return to using more cash simply because it's no fucking business of anybody but me if I want to enjoy a coffee with a friend, drop a few pounds into a collection box or Big Issue sellers lap, pick up some fresh produce from a kerbside stall or stick a tank of diesel in my son's beat up old Peugeot Partner.

[Nuncio]

Clearly, Proverbs 15: 3 is relevant here. "The eyes of the Lord are everywhere, keeping watch on the wicked and the good."

I hope Genesis 7:20 is not "The waters rose and covered the mountains to a depth of more than fifteen cubits"