Law enforcement intervention set (contains definition of facilitator)

mariopaolucci commented 5 years ago

(from UCSC, Niccolò)

This scenario presents alternative choices in repression targets law enforcement agencies (LEAs) can choose between in the PROTON ABM. In particular, the possible targets will consist of OC group bosses/lieutenants (in the model, potentially OC agents with high score in measures such as betweenness) and workers in "facilitator" positions. "Facilitators" include logistic workers, such as long-distance truck drivers and airport workers, and legal and financial advisors. These agents have increased opportunities for crime thanks to their work position (e.g. drug smuggling, money laundering). This scenario will put to the test and compare the effectiveness of LEAs targeting OC bosses/leaders and facilitators toward reducing crime rates and OC recruitment, both as separate and combined policies.

Targets:

1. Workers in facilitator positions

Policies:

oo** Law enforcement targeting:**
Higher scrutiny of facilitator positions (will decrease facilitators' probability of crime commission and OC tie creation)
Higher repression of criminal facilitators (will lead to higher imprisonment rates for criminal facilitators)

2. OC bosses/leaders:

Policies:

oo** Law enforcement targeting:**
Higher scrutiny of OC members (will decrease their ability to commit crimes and consequently create OC ties)
Higher repression of OC members (will lead to higher imprisonment rates for OC members)
Active OC group disruption (OC ties will be targeted and disrupted)

mariopaolucci commented 5 years ago

The current implementation might mix up groups 1 and 2

cecconifederico commented 5 years ago

Lavorando nell'implementazione dei facilitator sono tornato sulla questione employers. Gli employers vengono creati, sviluppano una loro rete ma non partecipano alla realizzazione di reti criminali. Ovvero non sono PERSONS? CORRETTO? Mi sembra di sì, quindi io dichiarerò alcune PERSONS come facilitator, solo tra coloro che hanno un lavoro

mariopaolucci commented 5 years ago

(UCSC):

ok with selecting facilitators at random. No need to use centrality or degree.
scrutiny: for this simply we run simulations with different number of faciliatators (if a policy could reduce the facilitators by 30%, then the result would be.."
repression: more likely to be arrested

danielevilone commented 5 years ago

Ok, thanx!

mariopaolucci commented 5 years ago

More questions about the facilitators with answers from Francesco:

should they be reached from the turtle that spawns the crime? Or they are reachable independently of network structure?

dovrebbe comunque esserci un meccanismo che controlla che il numero e la distribuzione per numero di autori dei crimini ad ogni tick sia corretta. Se si deve verificare un crimine con 5 membri e che richiede un facilitatore (non tutti i crimini devono richiedere il facilitatore), e in un caso non si trova il facilitatore allora il modello deve continuare a provare sinchè non trova un gruppo di cinque con il suo facilitatore...

note that this assumes a crime creation mechanism different from the one we currently have..

shall we count them as one of the crime members, or an additional one? If I am preparing a crime of size 10, is the facilitator the 11th or the 10th man?

direi che i facilitatori sono a tutti gli effetti crime members e non additional (rasoio di Occam).
should facilitators get tired? (normally agents can do as many crimes as wanted in one turn)

non vedo necessità di imporre limite ai crimini dei facilitatori (anche qui Occam)

mariopaolucci commented 5 years ago

Should be eventually solved by https://github.com/LABSS/PROTON-OC/pull/85

mariopaolucci commented 5 years ago

Things to do:

network intervention: double check the crimonology paper
- check baseline effect of arresting people on OC group. Importante: la polizia
- quanti OC members commettono crimini insieme? La baseline che tipo di intervento genera su OC members?
Introdurre - se sono 5 c'e' la possibilta' che uno non venga arrestato
Introdurre Un lag di sei mesi tra crimine ed arresto? Ok

verificare e ricordare che il numero complessivo di reati è legato al crimine originario

(for all this see issue #88

Gli interventi vanno anche pensati come limitati temporalmente - tipo, politiche di sei mesi

oo** Law enforcement targeting:**

Higher scrutiny of OC members (will decrease their ability to commit crimes and consequently create OC ties)

politiche di sei mesi impattare non su c ma sul fattore oc member (spegnere)

Higher repression of OC members (will lead to higher imprisonment rates for OC members)

probabilità di arresto

Active OC group disruption (OC ties will be targeted and disrupted)

targeting of nodes based on their position on the network??

mariopaolucci commented 5 years ago

Higher scrutiny of facilitator positions (will decrease facilitators' probability of crime commission and OC tie creation)

.... meno faciliatators dall'inizio

Higher repression of criminal facilitators (will lead to higher imprisonment rates for criminal facilitators)

arresto aumento del 50% (probabilità) (si', si estende a tutti) (toglie un po' di facilitators) (doppio effetto molto carino)

mariopaolucci commented 5 years ago

rinunciare e sostituire il crimine con un altro nel caso non si trovi il gruppo (per esempio come conseguenza dell'arresto dei facilitators) (provare N volte)

mariopaolucci commented 5 years ago

Active OC group disruption (OC ties will be targeted and disrupted) targeting of nodes based on their position on the network??

.. this is not yet clear to me. For now it's not implemented.

mariopaolucci commented 5 years ago

This is still missing

rinunciare e sostituire il crimine con un altro nel caso non si trovi il gruppo (per esempio come conseguenza dell'arresto dei facilitators) (provare N volte)

LABSS / PROTON-OC

Law enforcement intervention set (contains definition of facilitator) #60