ArcGIS Pro Intelligence - Link & Movement Analysis Demo

[Music] [Music] today we're going to focus on a fictitious investigation that is sort of loosely based around the 2009 vancouver gang war uh so first i'm just going to provide a little bit of background on the vancouver gang war so this particular spree of violence was instigated by the successful seizure of cocaine shipments um from mexican drug cartels by the mexican army so as a result of the limited cocaine supply the price of cocaine skyrocketed in vancouver making drug dealing a far more lucrative legal enterprise so there were several players in the vancouver drug scene including the united nations gang and the independent soldiers who traditionally acted as dealers for the hells angels drug suppliers but with cocaine trade becoming more lucrative that new players like the red scorpions began to imp encroach on traditional u.n territory the decisive moment came when the bacon brothers prominent u.n gangsters left the u.n and joined and later took control of the red scorpions together the bacon brothers and red scorpions were able to pose a significant threat to u.n dominance causing the eruption of a full-blown war so that's just a little bit of intro a little bit of background so we're going to jump right into arcgis pro intelligence right off the bat here so here we are we can see i've got it in the nice dark mode and we can see we have our nice tailored and simplified ribbons uh up at the top uh so the first thing i've done is i've just um you know all those gangs that i quickly mentioned i've just you know sort of digitized their logos and placed roughly where their headquarters are on the map so this is one you know nice way to visualize that particular aspect but um as you'll notice from my intro there's a lot of uh you know relationship and intrigue happening between these gangs and it's a little bit difficult to wrap your head around so um you know for me this was a great use for a link chart to sort of figure out you know who's working with who and who who's at war with whom so let's do that let's go ahead and we'll build a link chart right off the bat here uh there's two ways to do that the easiest way though is probably just to jump up here to our new analysis tab and then our first button up here is the link chart so we'll go ahead and open a new link chart so it creates a nice blank link chart for me and it you know handily tells me to just go add my first entity if you've worked with link charts before this is what we would call a node we just changed the name to entity so we'll add our first entity so i'm going to do is add that gangs layer that i've already turned on in the map and then i just need to give it a field that um you know i want to display it based off of so it's going to be name and we'll just keep this thing simple and just call the entity gang so i'll hit okay and this is going to bring all of those gangs in and if i have nice symbology in the map it'll bring that symbology into the link chart for me which um in this case makes it much easier to understand uh who we're talking about uh so now that we have that you'll notice up on our ribbon that our relationship type button has now been illuminated that's because we now have an entity to work with so we can start building some relationships or links so just link the gang entity to itself so gang to gang and we're interested in who's allied with whom so we'll just call this allies and then we just need to provide it some key fields so how we're going to link everything together um so in this case we'll grab the allied with and our target entities are going to be the name so we'll hit okay uh and just like that it's drawn the links on my link chart for me ready to go uh so i can already you know start to see who's allied with whom we'll add one more relationship type uh this one for who's rivaled with whom so we'll grab gang and gang again we'll call this rivals and then for our key fields going to be very similar we're just going to grab rival and then name again whoops grab name again there we go but okay all right so now we have our allies and our rivals in here now you're probably looking at this going well this isn't very clear of course because it's difficult to distinguish between the the two sets of lines um so we can come over here under the contents pane and as pierre was mentioning that's also new so we can come in we can go into our relationships uh and we can come in and look at this particular relationship allies and you notice when i select that my symbology becomes available so i can open that symbology and maybe i want to change that to a nice green color so there we go so we got our allied in green and for rivals we'll make that red okay so already that's a little bit easier to understand understand but i'm not too happy with the layout it's a little bit cluttered so i can come up here to my ribbon again and go ahead and change that layout uh let's go with organic see how that looks just gonna redraw all that and we get this nice you know it spreads everything out and we don't have any overlapping lines anymore so already pretty quickly we can uh you know for example look at the un gang and understand their strong relationship with the independent soldiers um and then they're you know sort of fraught relationship with the red scorpions and the bacon brothers we can also see who some of the main suppliers are you know of different uh you know cocaine shipments so that's a very very basic link chart and you know it really is a good visualization tool but we're not quite ready to do any sort of link analysis i think we need some more data we need to find some questions that we want to ask so another set of data that i have and you can see this in the contents pane is my gang members so i have information about different gang members so just going to open up that attribute table and what we have here uh is the name of the gang member if i know it i don't know the names of all of them i'll know their gang affiliation and gang affiliation here is crucial right because this is how we're going to link our gang members to our gangs we'll do it by the gang name so that's going to be our linking field and then the other thing we have here is just some extra information about each one of those gang members if it's available so for example if they have a vehicle associated with them that's been put into a field as well as the vehicle's color and we've also gone through and looked at their ethnicity as well so what i want to point out here is this particular table is a very good candidate for doing link charting and that's because it follows good database design principles and it has everything broken down into its smallest logical components so ins instead of having just a description field with all this information um in that description field it's all broken apart so if we had a description field like that there'd be a little bit of data cleaning we would need to do to to get a you know a very compelling link chart put together so uh all i want to do here is just do the same uh you know steps and just go through and add the different entities and build the relationships same process i just showed you so just in the interest of time we'll just skip ahead here and so all i've done is i've just added all of those gang members in and then i've added nodes for ethnicity i've added nodes for vehicles and i've added nodes for vehicle color so we have you know the same basic underlying link chart with our gang uh relationships underneath but then we have all our different gang members spread out here so we're getting close here this is as you can see a much more complicated link chart and potentially we could start doing some link analysis here but i think i'm going to add a little bit more data in and see what we get so what i want to be able to do is link my known gang members with some open cases that you know some that existed in 2009 and see if we can match suspects with known gang members so unfortunately that data um you know it's old from 2009 and all i have is it's in this old school powerpoint um not very useful right so it's you know i've got but it's it's hard for the computer to read but there's some good information in here so crucially you'll see down here that we have you know some similar uh fields such as uh you know the suspect's ethnicity if there was a vehicle involved in the crime we've reported that um and also the color of that vehicle so already you're probably seeing that we're going to be able to take that and add that into the link chart and then based on those links we should be able to narrow down some suspects so i have about 20 of these powerpoints now i could manually bring them into pro but that's pretty tedious especially you know 20 might be doable but 100 would be kind of unthinkable so luckily uh over here on the data tab i have access to locate xt which is an extension for pro and it allows me to bring in unstructured data so let's let's launch locate xd so here we are we're going to get ready to do extract locations uh pretty simple to use all i need to do is go browse to the folder where i have all that all those powerpoints so there's about 20 powerpoints uh in this folder here so i'll add that so that's what it's going to go search through and then i just need to give it an output feature class uh so we'll just call this open cases one uh and then we're going to do a little bit more tweaking just so that we can get that data into that you know nice database design principle and have those smallest logical parts so we'll jump over here into properties and is where the magic really happens if we go over here to extract attributes we can extract custom attributes and we do this by using what's called a gazetteer file so the gazetteer file basically tells locate xc what you're interested in pulling out of any type of document and then uh based on that it'll pull that data out and it will put it into a unique field for you so i've already built that gadget here for us just to save a little bit of time but let's just go look at it really quickly they're quite easy to build um so for example let's look at this case info so in this particular case what i'm asking locate xt to do is go find the case uh the keyword case number with a colon and any time it comes across that in any one of the documents what it's going to do is afterwards it's going to capture a set number of characters after that case number so that's perfect for something like case number which should be a standard set of characters so that's a good way so it's going to find case number as a keyword and then it's going to capture a set number of characters and put that into a field for me another example here is description so very similar it's going to find that description keyword but this case of course descriptions not gonna have a set number of characters um so i have a few options but i can tell locate xt you know what just capture everything until you hit a stop string and by a stop string we usually mean something like a hard return so if someone hit enter when they were typing um you know meaning they've moved to the next line uh that's when it's going to stop capturing so a bunch of different options those are just two so we've built that gazette here and essentially all we need to do is come up here uh to extract and we're just going to change this to two because i think i've already used the number one and we'll go ahead and extract that so it's going through it's reading through 20 powerpoints and actually it's already done so it just read through 20 powerpoints and it went and plotted all of that on the map for me i can also give it custom symbology so that it'll present that symbology right away and if i open up the attribute table of what we just generated it's got a lot of fields here about what it was doing and how it did the capture but crucially at the end here you'll see it's captured you know the ethnicity the vehicle vehicle color and then the case info and description so just like that we've been able to take you know sort of semi-structured data out of a powerpoint get it in darkjs pro and be able to use it in link charting and link analysis so let's um do that so we go back here again because of the demo i've cheated a little bit which is fair um i've gone ahead and i've just brought that open case data in and i've linked it all together because it takes the three or four minutes to build all those relationships up so we'll go up here to the diagram again so now that we've got our sort of i would say fairly complete link chart done with all the different data elements we want we can now start doing some data analysis so to do that we can come up here to our link analysis tools and we can open up and choose what sort of tool we want to run so today i'm going to focus on paths actually i'm going to focus mostly on neighborhood analysis we'll do a bit of path centrality and clustering very very useful tools uh they don't necessarily apply to this data set because um as you can see by this link chart i have lots and lots of links um so everything is sort of clustered together already this is good if we have you know lots of data and there's only one connection between each node or two connections centrality and clustering is really powerful um in that sort of case so let's start with our neighborhood analysis so the first thing i want to do this is that powerpoint we were looking at there that you know bc case 2009002 so i don't really remember what was going on there so i can just do a quick neighborhood analysis and just select on that node and just like that it's going to pull out you know the key things that it's so the things that are connected to it within one node so essentially the description of in this case the description of the suspect so we know that the suspect is caucasian drives a ford explorer and then that ford explorer is green uh now we could probably you know start looking at these links and see there's you know only a few links to ford explorer so we might be able to dive in that way but the other thing we could do is just extend our neighborhood analysis out to an additional node so by doing that it's now highlighted anything that's basically within two nodes of that case so what we have here is it's highlighted any one of the gang members that either are caucasian drive a ford explorer or drive a green car so it's basically generated a quick suspect list for us and so we're going to want to go ahead and narrow that down a little bit more but you can see how quick it is to you know take a lot of data and figure out you know what is interrelated most closely um so let's see here let's look at this don lyons i'm just going to pick him because he's just close in the link chart doesn't actually mean he's a close link but i just want to see what's going on with him so to do that i can switch from neighborhood analysis to a compare neighborhoods and for that i just need to grab my case and then i'll grab don lyons and it's basically going to highlight any of the nodes that are within one node of both of those so what this is telling us is that don lyons is caucasian and then he drives a ford explorer so you know that's pretty compelling stuff uh you know there's i think a high probability that um you know he might have been involved in this but i'm also noticing here that jamie bacon uh he also has uh looks like he also drives a ford explorer so let's do a quick analysis on him and see what we're going to get so same process we'll get grab this case and then we'll grab jamie bacon and you can see here that he actually matches even better so he's caucasian drives a ford explorer and we know his ford explorer is green um so we can narrow down on jamie bacon and see uh you know if if we want to put maybe you know more surveillance on him or see if we can see where you know if he's got an alibi for the time that this uh this crime took place so that's a quick neighborhood analysis i'm just going to show one quick other analysis so we'll erase this one and that's the path analysis um maybe not as practical here but for example if we wanted to know you know if we were looking at for example jamie bacon and maybe this guy up here barzan and we wanted to see you know how were they linked together we could do for example a shortest path analysis so we could select on jamie bacon and then select on barzan and just like that it's done a quick pass analysis for us so we can see that they are a minimum of four nodes apart from one another and that they would have to go through basically a gang that's at war with each other so you know if we were to see these two people interacting um we could say that you know that's that's a little strange they're not very closely associated so so that's just a quick uh path analysis that we could look at okay let's jump back over to the map now so here we have the map i'm just going to turn on the same data again so here we have this open case data that i ran just prior to this demo and what we're going to do here is we're going to create another type of chart which is called the timeline so to do that again we just come up to our analysis tab and right beside the link chart is timeline so this is going to allow me to break down all the different cases by the time that they were reported or that they occurred so to do that i just add my open case data and then i'm just going to display it based on the case info which is of course the case number so we'll hit apply so just like that it's generated a nice timeline for us so we can see you know quite a bit of activity happening in january and february of 2009 and then it's sort of petering out uh you know into the spring but the important thing about this timeline is it works interactively with all the other charts and maps that i've created so just going to give myself a little bit more screen real estate so i'm going to go grab our completed link chart put this on one side and then we'll just collapse down the contents pane and we can see now we get a nice view of all the data so we can view it spatially we can view it temporarily and we can view it based on attribute relationships the other thing too is if we come up to timeline it's also got its own ribbon but we can do a selection so maybe i'm interested in all the cases that happened in february and you'll notice it'll select all of those on the map it will also select them all in the link chart so it's a very nice way that we can you know look at our data in multiple different perspectives all at the same time and really get a better understanding of what we're looking at okay so we'll clear that and now we're going to sort of shift gears a little bit we're going to move away from this link charting and link analysis and move into the new movement analysis that pierre was talking to us about so we're just going to bring my screen back bring up my contents pane again and then bring up my catalog and we'll go ahead and switch here to the movement analysis okay so we're in movement analysis and what i have here is a series of gps points so we were lucky enough to get a court order for example and we were able to track uh you know several of the gang members over the course of a day uh so what we have here is just the gps points uh from all of those gang members over one day so let's open up that attribute table and see what we got so we have about 12 000 points and we can see that each of the gps points has a member name associated with it so that we can distinguish them and then crucially it also has a time stamp so we're going to need a time stamp to do a lot of this um you know uh movement analysis of course but of course if you look at this time stamp it's in unix or epoch time so basically seconds since 1970 which is a great machine readable format but pretty difficult for you and i to read and it's also not really the best design for using tools in arcgis pro luckily we provide a nice tool to do a conversion i know in past times i've had headaches trying to convert unix time so i just wanted to show everyone that it's quite easy to make that change so to do that i come up to the data tab and we come down here and there's this handy tool called convert time field so we'll open that one a very easy tool to use all we do is feed it that gang member tracks data and then we input the time field that we have which is called time and then we just need to give it the format of it uh and we can just simply tell it that you know what that's in unix oops uh yeah unix and underscore s for unix seconds and we're going to output it to a field called time converted and we're going to call that a date field so we'll hit run so that's going to go read through those 12 000 records or so and you'll see in the attribute table it just went ahead and added a nice human readable date time for each one of those points so once that's done we can now start diving into our analysis okay so the first thing i want to do is sort of make some sense of you know what i'm looking at here is just a big cluster of points not that easy so what i'm going to do is come back to our data tab here and we have this handy tool called points to track so we'll use that and all we're going to do is give it the input of those gang member tracks we're going to use that new date time field called time converter that we just created and then we're going to group it by that member name so that we have a unique track for each one of the members throughout the day and we'll go ahead and run that it takes about 30 seconds to run so i'm just going to skip over it and show you what the output would look like so here's our output so we can see we got this nice color coordinated output that shows us the tracks based on the gang member name the other thing that's uh really neat about it is if we open the attribute table it does a bunch of calculations for us as well one of them being speed so it will calculate um the distance and the time uh delta between each and every one of the points and then we'll output a speed value for you so uh that's useful just you know if speeds a factor in some of the data you're looking at so we'll close that guy up all right so now looking at this data there's sort of two main things that stand out to me one and i think this stands out the most is if i come up here to central city shopping center i can see that you know a lot of the tracks lead into this central city shopping center to me this sort of screams meeting location or drop point or something so i'm going to want to investigate that a little further the other thing that might not be as obvious but does jump out to me is this meeting looks like that happened between jonathan bacon and michael lee so they're actually on opposing gangs and i you know it'd be strange if they met like this so i would definitely want to know if that was the case um so i just want to run some analysis and see you know have the are these meetings that they go there at the same time stuff like that so to do that i'm going to come up to my analysis tab and we'll come down and we'll use this new tool called find meeting locations so we'll run that another fairly simple tool to use all we do is feed it those gang member tracks and then we just need to give it the name field which is again the gang member their name and then we give a criteria for what we consider a meeting to be so for me i'm just going to leave the defaults and for the defaults what that means is i in my definition of a meeting i think it happens within 100 meter area and that they need to be in the same 100 meter area for a minimum of 10 minutes before i consider that to be a meeting of course depending on your data you might want to change that and you're fully able to change that however you see fit but for me we'll just leave the defaults again because of 12 000 points it takes a little while to run so uh you know the magic of the demo i've pre-run that for us so we come in here we can see that it has actually identified this west oaks mall as a potential meeting area we turn this back on it you know aligns up nicely with you know those points that we had the other cool thing about the meeting location is again if i open the attribute table we can see that it gives us a meeting duration so how long they had met for and it also captures the time so uh you know they started to meet around 11 22 and their meeting ended 11 around 11 48 so fairly short meeting but it did meet our criteria uh so that's really uh that i think that's very valuable and we can say you know with fair amount of certainty that you know jonathan bacon and michael lee did in fact meet um but also what's interesting is what got left out so you'll notice that nothing happened here we did not get a meeting location so what that means is none of these tracks met that criteria so they were never in the same area within the same 100 meters within 10 minutes um but i still want to know what's going on here so for that i can use a different tool and that's this compare areas tool so to do that i just need one other piece of data i just need to digitize a quick polygon so i've already done that so this polygon i've just highlighted the the shopping center and i've just given it a name called central city shopping center so all i need to do is give me the give the gang member tracks the area is going to be this investigation area and then our point features the name field is going to be member name and then the area name so that's where i gave it that central city shopping center now i could do an analysis based on location and time that's probably going to be valuable but let's leave it at location for today and we would run that again i've already run that for us so uh when it finishes running we get this fairly boring looking output which just simply matches our compare our investigation areas polygon but if we open up the attribute table for compare areas this is where the interesting information is so it will create a list of all the different members who visited that you know were in that area and it also does a count of the gps points within that area so we get a quick sense of you know who spent the most time there so we can see pretty quickly uh the different individuals who went there so to me this sort of means we're missing something in our information so either this is considered a drop point and maybe we want to set up some surveillance or they're meeting with another individual who we are not aware of so you know maybe again surveillance would be a good idea to figure out what's actually happening in this in this parking lot all right so that's basically the conclusion of my analysis but i just want to say that you know all of this work that we do in pro intelligence is of course uh work it can you know work with the rest of the platform so if we come back to my story map here and scroll down we can see that that link chart that very basic one anyways that i created we can export that as a png they can also be exported as graph ml and xps i believe which are two different formats so potentially you could be able to bring those into something like i2 we can also you know any of those maps that we created and any of that data that we captured using locate xt we can bring that right up into the web and share it as part of our briefing tool and then of course any of that movement analysis all of that data can be converted into feature services and we can host it either on arcgis online or portal for arcgis you

Channel: Esri Canada

Views: 206

Rating: 5 out of 5

Keywords: ArcGIS Pro Intelligence, Link Analysis, Movement Analysis

Id: B2kyNSNKj20

Length: 28min 19sec (1699 seconds)

Published: Thu Apr 08 2021