This is a capstone project on the effect of transparency and controllability on user’s willingness to give permissions to Lufthansa. In this study we designed and implemented the Website mockups for Lufthansa ID registration precess and preference panels in different transparency levels. Then, we did a human subject research to test how the transparency of language during registration affects users' willingness to register for a new service. We also test how does a clear and understandable presentation of permissions and preferences influence the the number of users willing to opt-in.
TravelTrail is an Android App, which can track individuals traveling path, manage photos taken on the way and record these memorable sceneries or stories they come across during the trip.
This app takes use of Camera and MediaPlayer APIs to record when, where and what photo did user took. It can customize a map by setting image thumbnail as markers, cluster nearby markers as one marker and show trip path. It uses Parse APIs to support: user login/sign up and profile update; save, retrieve and manage markers.
JobDepot is an iOS application that connect job seekers and recruiters. It allows job seekers to search for jobs based on industry area, locations, and apply jobs from CareerBuilder Website. Recruiters can post jobs to our database and search for candidate based on candidate's background.
There are three third party APIs has been used: Parse iOS APIs, CareerBuilder and LinkedIn. Parse iOS API is used as remote database storage, as well as user authentication management. CareerBuilder API is used for search job and apply job in job seeker side. LinkedIn API is used for the candidate to import resume information.
Android Apps Collection
This app is a simulation to the website: http://www.mlcalc.com. It is used to calculate the expected morgage payment.
Manage student scores using SQLite
This small app is designed to keep records and perform statistical analysis for a class of students. It uses SQL Lite database to save and retrieve student record.
Vocal artist app
This app can be used for marketing a vocal artist. In this app, you can access this artist's songs, videos, images.
Large Scale Web Text Classification
In this project, we organized natural language data from World Wide Web and giving them class labels. In order to recognize to which class the data (i.e. Web page) belongs, we need to find a method that can make a good classification of these pages. The page data was denoted by bag-of-words method using both unigrams and bigrams extracted from head and body of the web pages. We developed two classifiers based on AdaBoost and Kernel SVM and then analyzed both results. Finally we found that the Kernel SVM runs faster and gets a better accuracy result, which could reach up to 95%.
Password Manager Usability Study
The purpose of this study is to determine what attributes of an in browser password manager improve its ability to support a typical user in the course of regular web browsing. The study includes a lab study to test the usability of different password managers (LastPass, Blur, PasswordBox and Chrome built-in) by individuals with varying technical background.
In order to preserve privacy of the website visitors, there are several web extensions, such as GhosteryGhostery, Lightbeam, Disconnect to help to manage the trackers. In this project, we analysised these extensions and combines the pros of these extensions and designs some key features that would surely help the users understand better what the trackers were doing with their data.
Grumber is a sophisticated nanoblogging site. I designed and implemented a featureful, interactive web application including user registration and authentication, email integration for user verification, photo upload and sharing, search and quasi-real-time updates.