This course module deals with human trafficking which is a serious problem in every part of the United States and across the globe. If you need help or want to get help for someone, reach out to the Human Trafficking Hotline at 1-888-373-7888. We've gone through some specific examples of how computational thinking is being applied to the domain of human trafficking. We're going to talk through a few more examples to clarify and give you some additional things to think about with respect to how computational thinking breaks down and is applied in this case. Let's look at another example of computational thinking at work on human trafficking. We're going to start off again with problem identification, which is the primary step when we do computational thinking. So here's a graphic to help you think about computational thinking in a particular flow chart way. Think about human trafficking as a problem on multiple levels and scales. Then we narrow that down and we also think about the places and routes. So human trafficking involves movement, and involves movement of people from place to place. We've already talked about this. So if we want to prevent and intervene in that movement, well we need to understand where things are happening. Understand the movement and geography of trafficking in order to do so. So let's focus in on a more specific example to really clarify what I'm talking about. Here's a problem statement. How can we map the routes of human trafficking so that we can better prevent it and intervene when necessary? How can we figure out where it starts, what's the midpoint, where it's ending, and where things are happening? How do we make sure we have the right resources in the right place at the right time to intervene in trafficking and help victims and survivors? Remember at the beginning of this case example, I asked about those posters and stickers that show up in rest stops. How did people decide where to put them? Are they in the most impactful place or other places where that information is needed more? So again, we're going to take this issue, this problem statement that we've identified, and we're going to put it through some decomposition. We're going to break it down. So we've got this broad concept of the movement and geography of human trafficking. Let's break that down into two components: the means of transport and the routes of transport. The means include the physical means of transport. We're talking about people being brought in on trucks. Are they being flown in? Are they moving through airports? How they actually getting from place to place. There's also the financial transactions connected to transport. Who's running the truck? Who's buying the gas? Who's paying for this? Who's making money off of it? Then we have the routes of transport. Where does this all begin? Where is somebody captured, abducted, tricked, brought into this horrible cycle? What are the routes? Where are they being moved from? Is that all staying within one city? Are they crossing state lines? Are they moving to different countries? What are the trafficking locations? Where is the actual trafficking taking place? Multiple locations? Is somebody stuck in a single location? Where's this happening? This data would be gathered from multiple sources. From human trafficking hotlines, from service organizations that help people, from law enforcement, from victim testimony, and any other sources where such information can be gathered. So now we've broken that down, we move into abstraction and we begin thinking about how that information can be turned into measurable and observable variables. What is the scope and scale of interest? Global, national, regional, what are we interested in, and what's our problem focus? What are the actual data points? How are we going to convert location into data? What information is needed, what information is not? Remember that in abstraction, it's also about getting rid of the information that we don't particularly need. So here are some possible variables of interest. We can convert location into GPS coordinates of all identified locations, and then we can code each location. Is it an origin point? Is at a place in a route of trafficking? Or is it a location where trafficking is actually taking place? All the GPS points can be coded as well for the type of location. Is it a hotel? Is it a rest stop? Is it a nightclub? Then we might eliminate some variables. For this particular analysis, we might decide we're not really focused on the means of transportation. We're really more concerned about where it's happening and what types of places. So once we've done that, and we move in a pattern recognition, and we're looking to figure out the patterns among places, locations, people, types of locations so that again, we can intervene and stop people from being victimized and help and rescue those people already stuck in the cycle. So what are patterns in past data? It would be one of the first things we'd want to look at. What do we already know? When we map data points, what geographic patterns emerge? Are there locations where this is happening more than other locations? Are there clusters in different regions? Are there commonly used routes? Are there pathways, trucking routes, highways? Are there patterns in the types of locations? Particular types of businesses that are frequent and are turned into trafficking locations more than others? So are there patterns then in the types of locations when they get cross-referenced with specific geographic locations? Again, thinking about, are there particular types of businesses in particular geographic locations that should be the focus of investigation? So when we thought about the patterns we might look for, and when we've investigated patterns that past researchers have found, then we can move into algorithmic thinking. How do we break these problems down into instructions and rules for our computer. What do we want the computer to look for? What is the specific problem we want to solve with data? This requires us to return back to our problem identification phase. So this is an iterative cyclical process. It doesn't necessarily happen step-by-step. We might need to go back and think through some of the other data that we've gathered. Really what we want to do is better predict where trafficking starts so that we can improve prevention efforts there. So how do we do that? Then we might frame this question. We're going to turn this into an algorithm for the computer. How do we better predict locations where human trafficking is occurring so that we can focus intervention and interdiction efforts in these specific locations? Where do we want to put our resources? So we have this question and that's going to help us begin to create an algorithm. You can think about an algorithm almost as a data table that's going to help us think about how we're going to operationalize variables and ask the computer to look for patterns. So this table represents algorithmic thinking that we might do. We might start off by thinking about the cases for which we have data, each case represents a survivor, a victim of human trafficking. For each person, there is a location of their original contact, and we can associate that with GPS coordinates and the type of location. We can also consider where they got out if there's a survivor, if there somebody we know about, they've probably escaped, they'd been released, they've been rescued, or somehow they entered into the system. So where did that happen? Have they had contact with law enforcement? Or have they had contact with any social service organizations? Where exactly were they trafficked? When they were forced into sexual exploitation or when they were forced to work, where did that take place? Other locations they visited? Rest stops, restaurants. what information can we gather? So for each of these types of data points, we're going to plot it on a map, and we're going to code it for the type of location it is. Then we're going to give the computer an algorithm to look for patterns of movement and trafficking across large numbers of cases. We can also add demographic information to the cases. We can add in age, gender, race, national origin, all different variables that we could enter in again to help the computer look for patterns to begin helping us target our services and resources to help people not ever get traffic to begin with or get out if they're already being trafficked, and then to hold those doing the trafficking accountable. So once we've applied our algorithm, we begin thinking about evaluating the data we've gathered, evaluating the solution set that we're developing. We're going to start off by actually using the algorithm to help guide law enforcement and prevention activity in specific locations. So then we have to gather data on enforcement success in places not using this approach. How many cases are they successfully prosecuting? How many people are they helping? We would want to compare that reinforcement success in another area where computational thinking is being employed to see is it actually working. Or there may be other means. But basically, we're going to apply computational thinking solutions, algorithms to help us figure out if this is actually making an impact. Are we making a dent in the trafficking activity in a particular area? So as we're beginning to wrap up, let's think about one additional problem, connected to the problems we've already been thinking about. How do we know where to concentrate the limited resources we have to help people who are impacted by trafficking? How could we leverage basic statistical analysis for example, to generate profiles of victims so that we can better understand who is at risk, when, and where they're at risk in order to develop more effective and accessible prevention and intervention efforts? For example, could we create an algorithm that scans data gathered in hospital intakes to identify potential victims? By the way, this is something that's already in place in some locations. To be thinking about a possible victim, they might have demographic characteristics that align with noted patterns connected to human trafficking, a particular profile of someone that is more typically victimized than others. The place and time of their entering into a hospital might also fit past noted patterns. There may be additional risk factors that seem present, particular age, immigration status. There's all things that we can find in the data that might suggest that a person's more vulnerable than somebody else. There may be other characteristic patterns present. Particular types of injuries or ways of talking about family or their daily routine that an intake worker might notice. So there may be multiple ways to potentially identify that someone seems more at risk of being victimized by human trafficking or might already be in that situation. By looking at this data, and gathering this profile, and then creating an algorithm to quickly map on new incoming patients to this, hospital workers can identify people who might need an additional intake questionnaire, or they might just need some services offered or a little bit of compassionate probing from a worker to see if everything is actually okay. So now you've had some time to think about a range of different examples of how computational thinking is being used as an important tool to help address the issue of human trafficking. We're going to close out with one more example that you can think about so that you can get some practice applying the concepts and principles of computational thinking. I'd like to encourage you as you learn about other issues of social justice and issues of complex social problems, to think about how computational thinking can be used there. Computational thinking is currently being used across the world to investigate human rights violations. To investigate mass incarceration. Here in the United States, computational thinking is being used to address issues like bullying and police violence in communities of color. So as you learn about these issues, about the pressing problems that we face today, think about how computers and computational thinking can help us gather the necessary data, look for patterns, and come up with new solutions.
