First of all, there will be challenges in collecting all these data without your bank, telco etc changing their EULAs and will be heavily infringing on your privacy.

Side note: The AU is coming up with an overly ambitious data protection policy that will even limit how and what apps you can use, they'll need licensing - the above proposal conflicts with this ( see my article which summarises the draft rules [I'm not a lawyer, may have missed or misunderstood some of the AU's good intentions] http://www.cio.co.ke/news/main-stories/african-countries-propose-stringent-rules-governing-ecommerce-and-data#)

Secondly, we don't have data scientists in Africa, IBM is repatriating(muffled laughter) some from the diaspora. Data science is more about people, than data.

The option here is to analyse the work of our police. But this would be leading the horse by the cat.

The police need computers first, and smart phones.

They then need to collect data, and good data, on potential suspects and suspect behaviour in a community.

The data then needs to be analysed and business intelligence done on it. Then, we may look at big data, coupled with other external data such as volume of calls in and out of the country and their source/terminating countries, airport arrivals and departures, unemployment, forex inflows and outflows, alshabaab twitter activity(yes, those accounts you kept shouting to be closed, alqaeda internet activity, global movement of alqaeda big wigs etc

In short, big data is the last shot here, and a safety can be up by a huge factor by implementing the rest without big data.

Lastly, we have a problem. Those in charge are already suggesting arming of civilians, which to me suggests admission of incapacity to handle the matter.

*Sips water from the bishop's well*