User:Fischmann

Project Description

Online crowdfunding platforms are popular sources for raising money for various causes. These causes can range from paying for sick animals' hospital bills to raising funds to cover a friends' college tuition. Often, online crowdfunding methods have come to represent the social and political climate, as those who are personally or tangentially affected by what they see in the news turn to crowdfunding as an opportunity to help others. With the recent crisis in Ukraine, many have looked to crowdfunding as a means of helping those across the world.

The objective of this project is to use web scraping techniques to retrieve data from projects related to Ukraine on crowdfunding websites. After doing so, the goal will be to determine through various methods what factors play a role in donations to the Ukraine crisis.

Work Log

Day	Description
6/1: Talk and Working on Script	Talk on Good Research Practices and Keeping Logs Spent time learning about and installing necessary packages Worked on web scraping script Worked on log
6/2: Working on Script and Mentor Meeting	Continued working on scraping data Met with mentor Worked on setting up environment
6/3: Working on Script	Continued working on script
6/4: RCR Training and Script	Attended RCR Training Continued working on script
6/7: Mentor Meeting and Script	Discussed progress with mentor Continued working on script
6/8: Script and Talk on Technical Writing	Attended talk about technical writing Continued working on script
6/9: Finalizing Script and Mentor Meeting	Met with mentor, discussed literature review, possible future analysis, and finalization of script Worked on last pieces of script, ran script on all data
6/10: Collecting additional data	Collected additional data that could not be collected via script
6/13: Cleaning Data	Cleaning Data
6/14: Cleaning Data and Meeting with Mentor	Continuing to clean data Met with mentor and discussed locating relevant literature
6/15: Finding Papers about ML/Crowdfunding and Research Talk	Research Presentation by Dr.Madiraju Worked on collecting and reading papers that used machine learning techniques to examine crowdfunding
6/16: Meeting with Mentor and Working on Data Visualizations	Discussed potential data visualizations and mini presentation with mentor Started to get familiar with Tableau, and worked on map making
6/17: Working on Data Visualizations	Started working in Jupyter Notebook, installed and began trying to make pairs plots Continued working on creating maps, figured out how to group data by country using a mix of Excel and Tableau
6/20: Working on Data Visualizations	Continued working on data visualizations in tableau and Jupyter Notebook Worked on fitting simple linear models
6/21: Working on Data Visualizations and Meeting with Mentor	Discussed improvements on data visualizations and different takeaways from visualizations with mentor Started editing visualizations
6/22: Talk from Dr. Bialkowski and Working on Data Visualizations	Research Presentation by Dr. Bialkowski Finished editing data visualizations
6/23: Meeting with Mentor and Data Cleaning	Started learning about Doc2Vec Discussed with mentor about future steps in project Started cleaning more data to be able to perform next steps
6/24: Data Cleaning	Worked on retrieving video transcripts Cleaned more of the data
6/27: Video Analysis and Doc2Vec	Looked at how to use Google Video Intelligence API/set up environment Started to implement Doc2Vec on data
6/28: Doc2Vec, Mentor Meeting, Presentation Prep	Looked at how to implement Doc2Vec Met with mentor and discussed presentation and future work Prepared for presentation
6/29: Mini Presentations and Data Cleaning	Practiced for presentation Presented research project and listened to other presentations Cleaned update section of the data
6/30: Sentiment Analysis and Mentor Meeting	Worked on running sentiment analysis on data Met with mentor regarding analysis and future work Continued working on sentiment analysis
7/1: Sentiment Analysis	Worked on combining data from different sources into one place Ran tests in R on emotions data