Jennifer from Data Science
Download MP3We know our services and programs impact you, so let's talk about them. Together, we'll learn from the people behind the scenes and get to know our city a little better. From the city of Winnipeg, this is our city, our podcast.
Natalie:This podcast is recorded in Treaty 1 territory, the home and traditional lands of the Anishinaabe, Enu, and Dakota peoples, and the national homeland of the Red River Metis. We acknowledge that our drinking water comes from Shoal Lake forty first Nation in Treaty 3 territory.
Tamara:Hi, I'm Tamara. Natalie, today we're talking about data. Wait, you say data. Data?
Tamara:Data.
Natalie:Oh, how do the difference?
Tamara:Yeah. Well, it's a good thing we have an expert on today's show about it. Yes.
Natalie:We have a data scientist. I I know that that's what she is. I'm just wondering about the pronunciation. We'll find out from her. We've got Jennifer Bodnarczuk joining us to, well, set us straight at least on pronunciation.
Tamara:Yeah. And find out more about our open data portal or data portal. Let's find out.
Natalie:Well, hi, Jennifer. Thanks for joining us. You know, walk us through your very unique path into data science.
Jennifer:Sure. I think data science is one of those fields where everyone has a unique story, but happy to share mine. I started in high school, actually, I worked in the local drugstore in a tiny little town in Minnesota, and I decided I wanna be a pharmacist because I really like what I'm doing. And so I went to school. I studied biology and chemistry and physics and calculus and, you know, was interested in it.
Jennifer:But then I took a psychology course, and I thought, wait a minute. This is what I wanna do. And so I changed my path slightly. I got a double major in biology and psychology. I moved to Winnipeg here, got a PhD in psychology, and realized my favorite part of learning about research psychology and doing experiments was the data analysis part.
Jennifer:And so I thought this is what I wanna do and I just so I've always like always done what I wanted to do at the time and it's really led me on a very interesting path and now here I am at the city to make a long story short.
Tamara:What is it about data that you love?
Jennifer:I think I am definitely an analytical thinker like I loved the when I when I started into psychology, it was like the statistics class that I really I'm like, oh, this is like the logic and the math and that, like, it just it was so much fun for me. And I'd like, twenty years ago, statistics was not everybody's favorite class. Whereas now, it's like data like, data science is popular and everyone wants to do it. So when I started in it, it certainly wasn't what everybody wanted to do, but it was what I wanted to do.
Natalie:Now to take it to today, overall, there are three big ways you and your team are working with data here at the city. This is at the project level, creating organizational tools, and then with just overall transparency. We want to touch on each one of these sections. Starting with, maybe you can give us an example from the project level. You've got a really great example to tell us about, a project that you did with the transportation division, and it's a case where you suggest sometimes there might be too many options for a human brain.
Jennifer:Yeah. The example from transportation is for a few years ago, certainly, but still one of the most interesting ones, I think, that we've done. So transportation had a few of their traffic counters out in the city, but they had more counters they wanted to place in new locations. And it's like infinite number of locations that these traffic counters could have been placed. So my team did a cluster analysis.
Jennifer:So we looked at the different types of patterns that were on the roads. There might be a road that's busy in the mornings, but not in the, you know, not as busy in the afternoons. So we looked at different patterns that we could see on the roads and then picked, like, the an a good example of that kind of road and then shared our results back with the transportation folks, then they made the final decision. So I think I I like this story. Many of my stories I think of the data analysis.
Jennifer:It's always a back and forth with us and the department. Certainly, have the knowledge of the techniques and how to use the data, but we aren't the subject matter experts on the actual topic. So we do a bit, we share it with the department, and then they make the final decision.
Tamara:Now something that folks might not know about is the city's high poverty area map. This is a tool that your team's involved with, and it's referenced across the organization as a consideration in decision making. Right?
Jennifer:Yeah. So the map of higher poverty areas was built around the 2016 census. It was actually a a group in community services that had originally developed the map. Then when the 2021 census came out, the person who had been most knowledgeable about the map had left the city. So my group was asked if we could update the map.
Jennifer:And so, certainly, we we said yes. We always say yes to to data projects. There's there's hardly anyone that we will turn down. So this one was a good example where we dug into what had been done in the past, how the map had been developed. It's actually all documented in an admin report.
Jennifer:So we dug back, found all the details of which datasets had been used and which variables had been used, and then we updated it using the 2021 census.
Natalie:And then the final way that your group is using data in a in a big way here at the city, is something that, you know, maybe citizens might not know about. They have more access than you might think when it comes to city data. A cornerstone being your work with the Winnipeg's Open Data Portal. Can you tell us about that one?
Jennifer:Yes. For sure. The, Open Data Portal first was launched in July 2014. And ever since then, we've just been adding more and more and more datasets to it. At the beginning of 2026, there were, I believe, 275 datasets and maps in total, and that's 215 unique datasets and then additional maps on top of that.
Jennifer:We always count the map or we've started to count the maps separately just, to answer various questions on the count of how many datasets there are. We're always looking for ways to make the open data more usable for residents and and for departments of the city.
Tamara:And we'll put a link to the portal in our show notes, but let's take a quick look through some of the titles. A lot of listeners might be familiar with three one one, but we can actually learn a lot from the three one one call wait times.
Jennifer:The 311 call wait times is a great example of there's a dataset on the open data portal. There's also the little ticker, the live counter on winnipeg.ca that tells you what the wait time currently is. And so that that little ticker count is connected right directly to the open data set. So you can look at historical wait times or you can look at what the current wait time is.
Tamara:So you're you're busy and you need to report that, you know, something needs to be, looked at by the city. Check the website first so you know really how long you'll be waiting. And our wait times are actually going down, which is great to see.
Jennifer:Yes. Yes.
Natalie:Another popular one on the Open Data, portal is, assessment parcels. I was kinda interested in what's going on there.
Jennifer:Yeah. I think that's one of our most popular datasets. I know when I've led student projects and things like that, it's a really interesting dataset for anyone new to Winnipeg because often they want to find a house. They wanna know about the neighborhoods and the assessment parcels dataset, especially for people who are data inclined to go grab the data, map it out themselves, look at all the different the prices of the houses and different neighborhoods and stuff. All of that information is in that dataset.
Natalie:So Every every property listed and value.
Jennifer:Yes. Yes. The tax assessed value is in there. So it's not exactly the market value, but the the, you know, the assessment value, which is pretty close on a on a grand scale. What do you think people are using that for?
Jennifer:I know that I think, like, real estate companies even will use the assessment parcels dataset to get information there because there's a lot of information about each of the each of the properties in that dataset. So it's a very popular one. People also are very interested about historical assessments which is something that is not in open data but it is something that, can be asked for like if if people contact open data, I can put them in contact with someone who can tell them more about historical assessment data or other data sets around the that the city has as well because that's another service that the Open Data team provides is is knowing about more than more than the just the data that's in the Open Data Portal, but where to find other data sets as well.
Natalie:And you're touching on this. So, you know, your interactions with citizens using data can sometimes result in your team, you know, working with departments of the city to to post more information.
Jennifer:Yes. There's a a recent example, actually. Someone a a resident emailed OpenData and said, I see that the open that the city has the roads and the sidewalks and the bike lanes, but what about the back alleys? So we reached out to the department and said, you know, the a resident has requested this dataset. Are you willing to share it on open data?
Jennifer:The department said, yes. And so now that's one of our newest datasets is the the the alley network and the map of the alleys.
Tamara:Sometimes publishing data is just the beginning. Tell us more about some of the interpretation included in some of the data sets.
Jennifer:Yeah. I think this is, something I'm personally very interested in because as a data scientist coming to open data, like, publishing the open data is, like, one step, one milestone, but then sometimes it's good to have another layer on top of that. I think a really good example well, it feels recent, but it was about a year ago. We, updated the way that the city was sharing the three one one performance stats. And so it used to be that, a team at the city would go into OpenData, create a new graph for the previous month, and then we would publish that graph to Open Data.
Jennifer:So we're building up this pile of graphs in the Open Data portal. But now if you go to look at the three eleven stats, it's an interactive web based, application that just points to the open data. So now any resident looking at winnipegs.ca now can see an interactive graph where they can choose the month that they want to look at and the graph will change right before their eyes and show them the data that they were looking for rather than searching through a list to find the month that you wanted. So I think this is an example of something that we can, as a as a city, move in the direction of adding more of these open data based applications onto winnipeg.ca that helps residents, media, others to use the data more than just look at a table of numbers.
Natalie:And then to really round this all out, we should kind of quickly clarify the information folks might ask for but but won't be on or isn't on Open Data and why.
Jennifer:We do get a lot of requests into the Open Data mailbox for data that we don't have at the city. A popular example, think, is the traffic collision data, which we don't have, but MPI, Manitoba Public Insurance, has that. Occasionally, people will ask about Statistics Canada data, which, of course, is a kind of open data, but it's published by the federal government, not by the city government. It's also notable that the police has a separate data source on their their own website.
Tamara:So the open data portal doesn't publish the police's data, although that's a popular dataset. Getting back to interpretation, you're someone who started with psychology but decided to stay with the numbers. Are there any common stories someone like you can see in spreadsheets?
Jennifer:This is an interesting question. I think it as I've worked my way through this career of looking at data in many different ways and many different datasets, think it's interesting, like, the textbook example that we're always taught is the normal curve, the normal distribution, which is higher in the middle and tails off in both directions. But in real life, it's really hard to find that distribution of data. And I'm maybe I'm getting too far into the weeds here of of data, but I more commonly in data, we find it's the extremes that produce the most visible outcomes. So this can be, like, a small number of people, you know, committing crimes, whereas the vast majority of the population don't commit crimes or the vast majority of people, you know, have relatively good health, but a few unhealthy people can can use a lot of the resources in a hospital, for example.
Jennifer:So that's always been fascinating to me that it's we think things are more normal. I don't know if that's like, certainly the normal distribution than they actually are in real life.
Natalie:Well and maybe to keep, you know, nerding out on on kind of the the grander scheme of all of this, another question for you. With more data and more AI, more artificial intelligence tools available than ever, what responsibility do public institutions have to ensure that the data we rely on is used fairly trend and transparently?
Jennifer:Yeah. That's a great question. I think with the popularity of AI and people, you know, to to some extent thinking that it can solve all these problems, but we still have to go back to the very foundations of, you know, statistics is that is the data that you're basing your analysis on, the data that the AI tool is using, is it representative of the thing that you're trying to get an answer about? And that is a really, really hard question to know the right answer to. And you and you always have to ask that question every time you're using AI.
Jennifer:Like, what is this AI basing its answer on and does that pile of information match what what I'm I'm looking at right now? So often, there'll be biases in data, you know, in in AI or in any dataset, really. It's always something that you need to consider.
Tamara:Now, Jennifer, this is something we asked The Economist in an episode last fall. If you could access one new data set, what would it be and why?
Jennifer:This is I know. I love talking about data. This is like every question I think is so interesting. So to analyze data, you need to the data has to have been collected at some point. And sometimes the data collection itself can be a struggle.
Jennifer:So often, data is produced almost as a byproduct of another operation system or something. So that's the data that's easily available and we can use it all the time. There's also another kind of data where it could be collected, but it would take an extra effort on top of the work that's already been done just to collect the data. So an example that I don't know why this example fascinates me, but maybe it's because I did see it in last year's, Datathon. But in the open data portal, we have the tonnage reports for the water and waste recycling and garbage collection.
Jennifer:But what we don't have is like the the weight of garbage produced by an individual household or even just a neighborhood level. Because to think and if you think about it, if that would mean that the trucks coming around to your house as they're picking up the garbage bin and dumping it into the truck, they would have to weigh that bin and associate it with the house that that is there. So that would be an extra step that is, like, currently just not, you know, feasible. So I but that would be one one example that I am quite interested in is, like, different neighborhood levels of garbage collection at and I have no good reason for being interested in this topic, but it's it's just kind of fascinating to me. And I think just a good example of data that could be collected and the different levels of which data can because we can collect it citywide, we can collect it at a neighborhood level, we can collect it at a household level, and each of those different levels will tell us different things.
Natalie:And maybe a last deep cut here in our discussion about data because you've shared in another conversation something that's really interesting that maybe ties into what you're talking about. The use of the use of synthetic data, where that can help when the real data doesn't exist. I'm talking about what I don't know, so so tell us a little bit more.
Jennifer:Yeah. No problem. And this is, again, on this the garbage and recycling data theme. This is what the so a team at last year's Datathon created household level data for the garbage and recycling and then made a map, this fascinating map of the different levels. And it was all made up data.
Jennifer:We understood that, but it was still interesting to see imagine what the patterns could be. And so often, even in in in, you know, psychological research as as my in my background or in other kinds of research, if you if you know what the distribution of the data would be or, you know, kind of how the larger level data would look, you can simulate or you can make up, create fake data that then fits those known distributions at a at a big level, and you can pretend that you know what the data is at the smaller level.
Natalie:To me, this kind of almost sounds like, you know, you rent a new place, you buy a new house, and you don't know until you get in there how to decorate things. You know, you don't know until you're you're maybe on the the back end or the or the outcome other questions you might have. And that's why that synthetic data might be helpful.
Jennifer:Yeah. Yeah. Think that's a good comparison where it's you can certainly learn different things from the different levels of the data. And, yeah, you don't know what you don't know until you've looked some of the data and go, oh, wait a second. What about this?
Jennifer:And then you can sometimes kind of work backwards and fill in the gaps where you don't know.
Tamara:Let's switch gears a little to something you've hinted at and mentioned a couple times already. Coming up this month is an event that really does tie this whole conversation together. What's the Datathon?
Jennifer:The Datathon is an event that we've tried to hold every year on Open Data Day, which is the first Saturday in March. And so we had missed a few years after COVID, but we had we had one last year. We're planning this year's Datathon. So it's coming up this Saturday, March 7, and it's really just an opportunity for folks. It's free to come and see what people can do with the open data.
Jennifer:So at the data time, you can participate as an observer or as a participant. So the participants will be competing against each other to see who can come up with the best project. And so the observers are there to to observe and to see what the, what the participants are creating based on the open data. So it's an all day event. We we have food, and we have lots of fun.
Jennifer:It's just I love it. It's so fascinating to see the different projects that come up. In the morning, there's different presentations, so we'll do a short presentation on what the open data portal is, how how people can use it. And specifically for for these data enthusiasts, how to use the the API to connect to the data so it's not going in and downloading an Excel file. They can connect with this API.
Jennifer:I'm sorry.
Tamara:I'm gonna pause you there. What is API?
Jennifer:API stands for application programming interface, but it's one of those acronyms that it almost people almost forget what it what it stands for. Yeah. But it it essentially, it's just a way to connect to a database. And so the data can flow automatically instead of having to manually go through steps to gather the data.
Natalie:You're also touching on, maybe gatherings of people I didn't even know were out there. These data enthusiasts. Who are they?
Jennifer:Sure. There many of them are students, I think. But there's also there's you'll find them every like, in all kinds of different jobs. We have a group on meetup.com. So there's there's a handful of data groups actually out in the city.
Jennifer:One of the biggest ones is just the data science meetup that we I do have a meetup group specifically for the open data datathons. And you yeah. You see a lot of familiar faces at these events, but a lot of, people coming for the first time that they've they've heard about this thing called the open data portal, and they want to learn more about it. So, Jennifer, when
Tamara:you look ahead five, ten years, what's the one way you hope data will make life better for you, me, everybody in Winnipeg that maybe we're not quite doing yet?
Jennifer:I think the one big challenge that we have ahead of us is actually for governments to share their data better with the other levels of government. Because I know I know the city has an open data portal. The province has an open data portal. We've tried in the past to kind of cross post some of the datasets, but it there's so much more that could be done. So yeah.
Jennifer:In five or ten years, I feel like that's a a realistic timeline where where we could do this better. And for the benefit of Winnipeggers, especially where in in the province of Manitoba, like so much of the Manitoba population lives in the city of Winnipeg. So a lot of the data that's relevant for Winnipeggers is also relevant for Manitobans and vice versa. And and then to recognize, you know, what what data are we missing? What data is out there in the rest of the province too?
Jennifer:It's just so, again, it's an kind of an endless endless search for more and better data and communication about who needs what data to answer which questions.
Natalie:Well, thank you so much. This has been so much more interesting than a spreadsheet. And what maybe, hopefully, folks, you've you've changed their mind and what they thought they were getting into today. Last question for you, and it's one that we ask everyone. It doesn't have to be data based.
Natalie:This is just something you wish that you could share, that that you wish everyone knew about Winnipeg.
Jennifer:So I think this is so I'll veer away from data here because one of my interests outside of data is actually compost ing, which may seem like an odd combination.
Tamara:But your answer is to the good question now. We're figuring you out.
Jennifer:It does. I know. Yeah. Yeah. There's pattern here.
Jennifer:You're recognizing the pattern. Yes. Absolutely. And maybe that's yeah. Good point.
Jennifer:That maybe that's my fascination because the more you compost, the more it reduces your garbage. Yeah. So then if we knew which areas of the city had more garbage, we could target our composting strategies and tactics to to those areas of the city. But I'm a backyard composter. I'll I love I've been doing I don't know.
Jennifer:I've lost count, like ten years at least. And just recently, I've I've completed the master composting training from Green Action Center, and I'm working on getting some volunteer hours to to complete that that certificate. But it's, it's just fast and it's so fascinating to see, like, nature working and you can throw in, like, these, you know, kitchen scraps and these leaves and just watch and pretty soon it just turns into like, they call it black gold, this this compost that you can spread then on your plants. And there's so many ways to do composting, which I think is that's what I would like to that I wish people in Winnipeg knew all the options that there are. So you can do it in your backyard.
Jennifer:I'm actually a seasonal customer of Compost Winnipeg, so they come every week right now in the winter because my my bin is frozen. But they'll come every week to pick up my kitchen scraps. The city has the green bins around at a few of the For our deepest Yeah. Yeah. And the community centers.
Jennifer:Exactly. And, you know, if if you find out your neighbors compost, you can give them your kitchen scraps. Like, it's a or and there's always the, if you're in an apartment, you can get the little red wiggler worms that that you just feed them your kitchen scraps and they'll eat them and then produce a compost material too that you can put on your house plants. So that's my one of my next goals. I wanna get some of those red wiggler worms.
Jennifer:I haven't done it yet,
Tamara:but If you ever get bored of data, I think water and waste might wanna kick your brain about a couple things. I have one quick question though and this is maybe just semantics. Data, data, we've been saying As it a few different
Jennifer:far as I know either way is fine. I know I myself say it different ways and yeah, I I don't know if I I don't know which one is the right way either. It's just one of those words that you can say either way.
Natalie:Thank you. Oh my gosh. Thanks so much.
Tamara:Oh, that last bit about composting makes me just miss warmer weather.
Natalie:Yeah. No. And, we have a great conversation, really to to grow into the next month because we are talking at the Living Prairie Museum,
Tamara:about Which is more than a museum.
Natalie:Oh my gosh. It is really about, you know, life and our our relationship. It's it's it's a theme. The prairies come up in so many of these conversations. Yeah.
Natalie:We're gonna go take a trip out there. And, you know, in in the in between time, if anything comes up from you, listener, you know, if there's something you want us to talk about, talk to, we'd love to hear from you. Send us an email over at city podcastwinnepeg. Ca with topics you'd like to hear about.
Tamara:Chat with you soon. Bye.
