Although the internet provides a wide breadth of information in an extremely easy manner, a common issue people face -- regardless of the type of information -- is an overabundance of information. Services like Yelp, TripAdvisor, and GoGoBot provide lots of information, but do not always provide an easy way to whittle restaurant/activity choices to help a user make a decision. Moreover, the results of a search may not always be relevant to that user’s personal interests, due to how general the results are. Curato attempts to provide a single, convenient application to help users find businesses and points of interest relevant to their interests by taking advantage of simple machine learning algorithms.

We are working with the School of Environment and Forest Science (SEFS) at University of Washington towards the mission of making environmental data accessible to Native American people. The researchers at SEFS have collected large data sets related to land cover, hydrology and precipitation to learn carbon emissions and absorption in that area. The challenge that they currently face is to make this scientific data available to the Native Americans in a format that they can relate with and interact. In order to facilitate this, we have integrated the existing environmental and ecological data and created data visualizations as a POC to ascertain that this process can be automated with the real time data. We have also prototyped user-friendly dashboard to present this information to the indigenous people in an intuitive way so that they can evaluate the health of their land and make better decisions about their environment.

In the realm of online content retrieval, understanding user intentions based on search behaviors is essential to connecting customers to the content they seek. Adobe Systems recently acquired Fotolia, a digital asset supplier, to more efficiently provide their customers with quality images and videos from inside their ecosystem. Working with Adobe Stock, we accessed all of their query and content engagement data collected to date. In this formative analysis, we applied natural language processing to search terms. Results were paired with multiple metrics of user interaction associated with Adobe’s content. These data were grouped, or clustered, to reveal hidden layers of similarity across queries. These clusters, representing similar user intentions, help identify Adobe’s underperforming segments of customer interests and behaviors. Once identified, treatments such as user interface modifications, search algorithm changes, or query refinement suggestions can be targeted to queries in the same cluster. This will enhance the Adobe Stock user experience¬ increasing customer retention, satisfaction, and spending.

Starbucks’ information security team is continually seeking to understand which of its security events to prioritize for response. Although the organization utilizes a security information and event management tool for detecting anomalous activity, the number of alerts being generated by the tool are overwhelming and difficult to manage. This is an issue that security teams at many large organizations face; how do you sift through the noise and find the events that are most likely indicative of a security threat or breach? IntelligentIR helps answer this question through the use of machine learning techniques. Using unsupervised learning to label raw security data along with supervised learning to build decision models, intellingentIR identifies and prioritizes new security alerts in order to make incident response more manageable.

Every admission season, schools struggle with attracting the best talent and achieve maximum conversion rate at the same time. They want to avoid both under and over filling of the classrooms. Therefore, there is a need to strategically balance the quality and quantity of incoming student population. We extracted and analyzed important features of the data provided by our sponsor Ravenna Solutions. Our predictive analysis shows how significant select factors are in determining a future admit offer or an admit conversion. Thus, our project provides key insights to admission directors of K-12 schools regarding factors influential in making admission decisions. We help admission directors in making crucial admission decisions backed by application data of student aspirants. This makes the directors rely less heavily of their instinct or “gut feeling” and instead make data driven decisions. Hence, the efficiency in decision making results in better enrollments for the schools and thereby better school admissions for the students alike.

PainlessVR is a research tool that allows researchers from medical, psychological and any other fields to test hypotheses about Virtual Reality without any development at all. The goal of the product is to improve future virtual reality pain management applications by removing the barriers to researching virtual reality. PainlessVR allows researchers to modify, and by extension study, variables regarding color, sound, context and cognitive load. Improving VR pain management would mean improving care for those who are unable to take traditional pain medication such as burn victims or recovering opioid addicts.

Patents are arguably one of the most important means of rewarding innovation and creativity. According to the U.S. Commerce Department, in 2010, IP-intensive industries accounted for 34.8 percent ($5.06 trillion) of U.S. gross domestic product. Despite this significant contribution, the process of acquiring and maintaining a patent remains fraught with complexities. Through our research project we intend to shed light on some of these complexities and provide data driven insights into this process. By scraping, cleaning, and analyzing 10 years worth of publicly available utility patent data, we have attempted to examine and visualize some interesting and pressing topics like prevalence of any gender bias, trends around industries/organizations that produce patents, and countries that spearhead innovation across the world, among others. We believe the data and insights produced by this project can guide future research and improvement efforts in this field thereby benefiting the patent industry, in a broader sense.

Each year, roughly 30% of first-year students at baccalaureate institutions do not return for their second year and over $9 billion is spent educating these students. Yet, little quantitative research has analyzed the causes and possible remedies for student attrition. Here, we describe initial efforts to model student dropout using the largest known dataset on higher education attrition, which tracks over 32,500 students' demographics and transcript records at one of the nation's largest public universities. Using a balanced dataset, an accuracy of 16% over baseline (66%) can be achieved. Logistic regression, random forest, and k-nearest neighbors models were used. Accuracy was boosted through a feature engineering approach. This project will inspire universities to use machine learning to identify at risk students. They can then target retention efforts towards these students; hopefully putting loan money to better use and ensuring that the most students possible earn degrees.

The company Test Innovators helps students prepare for standardized tests like ISEE/SSAT. Students sign up to their platform and take practice exams. They have more than 20,000 students signed up till date and pocess more than 5,000,000 questions on their system. The company wanted to build a system which automatically analyzes the student's progress in their practice tests and recommend ways to improve their scores. We have built a system, PrepSmart, that automatically learns from past experiences of students who are similar to the current test taker and suggest their features to the current test taker to improve his scores. Recommendations can be anything like strong/weak subject areas, time to answer questions,additional question banks and tutoring services. PrepSmart helps students in improving their overall scores and it also helps the sponsor in generating additional revenue by selling recommended question banks and tutoring services.