Roonak Rezvani is a PhD student from University of Surrey funded in part by the IoTCrawler and the European Union’s Horizon 2020 research and innovation program. She started her thesis in April 2019.
The title of her thesis is: “Timeseries- and sequence data analysis”.
Interview with Roonak Rezvani
We look at data in time frames and periods to see how the data develops over time. From the change in the data we can extract meaningful insights and which help us in predicting certain future events in relation eg. to traffic data and weather data. If you look at the time frame windows a lot can be said of traffic considering the month you are in, what time of day it is and what the weather is like. It can give you some knowledge of when to travel or commute next. If governments understand the data they can take action; travel time can be reduced, traffic flows can be planned differently and more efficiently and policies can be made to answer these newly discovered needs. As a private citizen you can also use this information in services for planning your vacation or commute.
Why did you choose this topic for your thesis?
My background is in computer science, I like to do programming. But besides that I want to focus my work on real-world problems and I want my work to matter to real people. My field of research and the way we work is closely connected to actual cases, so I can see the results and the product of my work and how it can affect people’s lives and I like that. Also I want my work to be challenging. Right now, I work with medical data and clinical data from another project. We look at interviews with patients, so that help us get more information out of the numerical data that we also have. It is programming as well as research. In research we work on creating and improving the methodologies on how to analyse these data.
How much of your thesis is related to the IoTCrawler?
My thesis is 50-50 related to the IoTCrawler I would say, but my work is not sharply divided. It can be related to the IoTCrawler. What I do with the medical data also inspires my research on IoTCrawler, and it can be used in the IoTCrawler research.
What excites you about your research?
I’d say working with numbers and mathematics which I like and I love how challenging it is. Every day I face a new challenge and try to solve it. I can see the outcome of my work and I see how the work I do improve the lives of others.
If you were to high-light one of your article contributions related to the IoTCrawler, which would it be?
I was part of an article on how to classify sequences of data. I looked at how the sequences of traffic goes – and we could classify a specific sequence of data – when is the amount of traffic high and when is it low. That was kind of novel, because before people looked at data in a different way and not as a sequence the way I looked at it. That’s important because sensor data, and the information from it, are dependent on time.
When you have big data – a big amount of data over several months – no one individual can learn anything meaningful from it by looking at it. Discovering patterns and gathering insights from this much data, stemming from different kinds of sensors is simply not possible for the human eye.
Traffic data is not just one attribute and traffic sensors measure different things. I take the data and represent them in sequences and put them into a trained model (the model was trained using the sequences of history data and event) and it detects the events from them. When you look at the data in the time sequences it is possible to get some useful and important insights.
What are your aspirations and hopes for the IoTCrawler?
The IoTCrawler will provide me and other researchers with a Google for sensor data. Sensors are everywhere, but it is difficult to get all the information, and all the data from them. From a research point of view, I would have a lot more material to work with and to have it all in the same place and to have it that accessible would be very valuable.
My advice for cities would be to make more data public. The city of Aarhus already has a wonderful open data service and the data there is very nice. The website is so informative and easy to work with. If more cities could do this, a lot of knowledge could be extracted from this by researchers. It could create value back to the cities and the researchers could work more efficiently. There are sometimes a lot of jurisdictions and decisions to be made before certain data can be accessed, which is holding back important discoveries and improved work flows.