About Me

I am a product manager in a start-up company focused on image and video technologies. In product development, data is at the core; meaning that, to determine how product behaves and to predict what is next for the product, PM needs to track meaningful metrics, analyze them and then utilize the insights to improve KPI for the product. My future plan is to become a data product manager; A PM who is working with data products need to understand statistics, machine learning and AI algorithms to some extent in order to build healthy data strategy- such as the plan for how data will be generated, collected, and consumed, and how this uniquely positions the product to win in the market. That is why i am here:) Beside this, I love storytelling with data. In my spare time, I search and find good data visualization works to ‘contemplate’ and inspire. This is another goal of mine :)

useR! 2019 Toulouse - Talk Social Science Marketing & business - Frédéric Vanwindekens

https://www.youtube.com/watch?v=gCbL3Sd-jKI

Vanwindeken’s research focused on visualization of qualitative data from open-ended interviews through qualitative coding and cognitive mapping. Qualitative interviews that are used in the study are in the context of social ecological systems like agriculture forest management, fisheries and etc. These systems, managed by humans and practices (the way of applying these techniques) are somehow different than the techniques. To understand the practice better Vanwindeken says that researchers should understand the worldviews/opinions of people. Qualitative data is gathered first form face-to-face interviews includes open-ending questions. Qualitative codings are main social science analysis method. Although quantitative methods are more common in data science with tables numbers and etc., this research aims to explore qualitative data by dissolving them into qualitative codings. Qualitative coding technique is attributing the codes to a part of a document. On the other hand, cognitive maps refers to a conceptual apparatus by which people classify, order and interpret the world. package for analyzing and treating qualitative interviews. By using cognitive maps, qualitative data(practices) is connected to quantitative data (techniques). In his research, qualitative data is turned into graph and those graphs turned into matrices. Then, matrices are formed into a social map that shows most important concepts and/or relationships in these systems.

useR! 2019 Toulouse - Talk Visualisation - Ursula Laa

https://www.youtube.com/watch?v=oiu9MwQ3BOA

Presenter’s, Dr. Ursala Laa, background was in particle physics and during the presentation she talks about how she used data visualisation packages to get better understandings of her physics experiments taken part in highly complex environment. She opens the presentation by showing an example from proton-proton collision from CMS experiment. In this data problem, the experiment has data set in high dimensions (means that high no. of independent variables in the data) and she wanted to understand concepts like grouping and outliers by using data visualization. Grouping in this case means that how different type of measurements constrain different directions in parameter space and outliers helps them to understand sensitivity (the experiment is run to understand sensivity to fit in single measurements). Grouping the data set with the help of data visualization (Grand Tour animation from tourr package) helps them to define the difference between groups and determine the multivariate outliers. She adds that it would be much easier to determine these if they add some interactivity on the results; therefore, she continues with galahr package (a GUI to assist learning high dimensions in R) that they used for interactivity.