Topictisat is a project based on R Shiny–Text Mining, Wordclous, Biagram Methodology
In this web application developed with R Shiny, the Twitter data of 102 economists is being analyzed. The aim is to reveal the difference in economists’ perceptions of the economic agenda by making two classifications: Heterodox and Mainstream economists.
Topictisat automatically writes the weekly data to the server every Monday. The data extraction process, which has been ongoing since January 2022, will come to an end in December 2022, and an empirical study will be conducted for the year 2022. Topictisat is actively working and continuing its development process at the following link.
Furthermore, the Topictisat project has been presented as a paper at the Why R Turkey 2022 conference. The Why R Turkey 2022 committee selected the project presentation as the best presentation and rewarded it with a $250 worth book gift.
R package used in this project: shiny, plumber, plyr, mfx, forcats, dplyr, ggplot2, plotly, etc.
Spectra AI – R based Churn Data Analysis– Logistic regression, bar & pie graphs
In this project, data from a security company with users in 81 provinces has been analyzed using logistic regression technique. Based on the calculated coefficients, the likelihood of customers leaving the company has been determined, and customers at risk of canceling their contracts with the company have been identified.
The Spectra AI project calculates real-time customer risk scores using R Shiny. Every new customer is automatically included in the data for risk assessment
R packages used in this project: shiny, plumber, plyr, mfx, forcats, dplyr, ggplot2, plotly, etc.
Covid19 Shiny Dashboard
Covid19 Shiny Dashboard – Leaflet project, map visualization, web data reading, etc.
In this study, global real-time Covid-19 data is visualized. Death and case numbers are mapped using the Leaflet package to display the data on a map. The web application allows users to compare Covid-19 data for different countries, providing the opportunity to visualize the case and death rates in each country.
R packages used in this project: Shiny, leaflet, ggplot2, lubridate, janitor, gdata, httr, etc.
This project is the task requested from me during the successful job application process at Appsilon, an international company based in Poland. In the project, GBIF datasets of over 20 GB have been selected and visualized, scaling the data for selected countries.
R packages used in this project: leaflet, DT, plotlyr, dplyr
R Shiny Text Mining
R Shiny Data Mining Project– Biagram Methodology, Frequency Bar Plots, Wordcloud, Biagram
In this study conducted for Istanbul Metropolitan Municipality, the data of 16 million Istanbul residents has been analyzed from the “Beyaz Masa” (White Table) system.
The web application developed using the Shiny package analyzes the complaints, suggestions, and demands of Istanbul residents at the district level in real-time. The study brings to light the complaints, suggestions, and demands of various societal groups such as public transportation passengers, workers, retirees, etc. by analyzing the data in real-time. Frequently used words, word groups, and word networks are visualized in the application.
R packages used in this project: shiny, forcat, igraph, ggraphy, plotly, dplyr, snowballc, quanteda, plumber, etc.