To obtain a full-time software engineer position in a high-quality engineering environment where my 4+ years' experience with interactive information retrieval system and full-stack web application design will add value to organizational operations.
Master of Science, Computer Science (GPA: 3.6/4.0) 05/2018 Johns Hopkins University - Baltimore, Maryland, the United States Bachelor of Science, Management Information System (GPA: 3.7/4.0) 07/2016 Peking University - Beijing, China
Bachelor of Science, Economics (GPA: 3.0/4.0) 07/2016 Peking University - Beijing, China
Find University: A website for pre-college students to search and discover target university Sep 2016 to Dec 2016 Built a web spider to crawl university-related data from diverse websites (Wikipedia, Times, U.S. News, U.S. Climate Data) using Scrapy and Beautiful Soup, and reduced garbage data by 95% by using Xpath to handle data extraction from malformed source pages
Organized and merged raw data by attribute (ranking, geographic, admission data) using Python, stored and managed data using MySQL
Developed a web application based on PHP that provides users with comprehensive details about different universities to browse and allows users to search for universities that match customized attributes (weather, ranking, expenses, programs, etc.) and enhance search experience by further filtering and ordering A clickbait detector for news materials based on external features Jan 2017 to May 2017 In this experimental project aiming to build a precise detector, our team experimented with multiple approaches in each phase during clickbait detector generation: vector construction, vector improvements and classification algorithm
Extracted external features (titles, keywords, snippets, etc.) from raw news data Constructed document vectors for each piece of news using two approaches: one is regular Vector Space Model using term frequencies, the other is to continuously compute vector representations of words based on bag-of-words and skip-gram architecture using Word2vec developed by Google Implemented two different extensions (bigram, SVD) to the basic document vectors using scikit-learn modules Implemented three different classifiers: cosine similarity to centroids, Naive Bayes classifier and SVM classifier Improved correctness from 46.4% to 74.4% after choosing the combination of VSM, SVD, unigram and Naive Bayes iHospital: A health helper offering doctor recommendation and appointment scheduling Sep 2017 to Dec 2017 Used GitHub to manage version control, cooperate with other team members, update design documents, track issues and plan for future features
Research Assistant Jul 2013 to Jul 2016
Peking University - Beijing
Leader of 8-people team dedicated to project "Research on Customized Information Retrieval System Based on Users' Search Behavior and Context"
Increased experiment data richness by 33% by implementing and configuring new software (such as Morae, cognitive style test, customized Firefox extensions, eye tracking devices) to record users' search behavior, collect any user-generated data, and track the pattern of their information search behavior Cleaned and reformatted raw data using R, analyzed data and plotting using SPSS Continually established clear operational and software requirements, assessed experiment results and re-designed methodology accordingly, which helped our team discover 2 more significant correlations between variables
Attended ASIS&T conference in Seattle, and presented our work to other scholars as primary contributor for 2 conference posters and 1 full journal paper