STORY 1

This is about my summer internship at Microsoft in China in 2020. I worked as a full-stack engineer in a team of 15 at Azure Department. Azure Cognitive Services was not very popular in China at that moment, and our team was asked to build some great applications based on ACS and show them to the public. My responsibility was to build a dictating and translating website from none. By using the speech to text conversion, translating, text abstract extracting, and emotion analysis APIs provided by ACS, as well as HTML+CSS, JavaScript and Python, I built the website based on Flask framework. The presentation of my website was edited into an article with gifs, which was viewed by more than 100,000 readers and led to a more than 5% increase in use of ACS. And the website was also awarded as the most powerful applications of the whole team in 2020.

STORY 2

Last April to June I attended an online CIS research. In this research students were split into groups for a free topic research. Each group contains 3 to 4 students. I was chosen to be our group leader for managing work and we intended to build a “steam games sales predictor” based on machine learning and natural language processing and write a paper. I took the responsibility to crawl game sales data and build the dataset. By using Python, I also built some models by applying Lasso linear regression, SVM, decision tree and random forest on sales data for the comparisons. Finally, by applying the SVM algorithm for the sales data combined with the applying of the random forest on game reviews, we made a predict model for the most popular 10 games in July. The prediction’s accuracy is 90%.

STORY 3

TFrom September in 2020 to May in 2021, I was engaged in doing my bachelor’s final project. After discussing with my mentor, I decided to make a research on the news about China in the biggest German magazine, The Mirror. This was an interdisciplinary project combined with German literatue, Linguistics and computer science. In the part of computer science, by using Scrapy framework in Python, I simulated the login process and the keywords sifting, I crawled all articles about China in printed The Mirror magazine from 1947 the beginning date of the magazine to 2020 December, which were more than 20,000 articles. I stored and managed the data in MongoDB. Finally, I analyzed the articles with many modules in Python such as Pandas and Spacy. Besides, I took advantage of another modules in Python like Seaborn and WordCloud to make some visualizations in my final paper. Finally, my final project received the highest grade in German Department and was awarded as the Outstanding Final Bachelor’s Project Design Prize in my university.