This project was an attempt to develop a thesaurus of terms to describe the mechanics of various game-based media. This process primarily consisted of seeking feedback from various stakeholder groups on what kind of language they used to describe these media mechanics, and what developments they felt would be most useful. Ultimately, stakeholder needs could not be accommodated in a single document, necessitating the construction and testing of multiple taxonomy structures.

Emergency youth shelters provide a safe, stable, temporary living alternative to youth who are dealing with family crisis. The staff at these shelters have the difficult task of balancing client care with meeting state and federal grant requirements. Our project saves the staff of these shelters time by tracking the most commonly needed client information and providing an online form database that can be updated when requirements change. With our tool, staff can keep an up-to-date list of the status of each case, allowing them to complete state and federal reporting while freeing up more time for client care.

Everyday millions of frequent flyer miles expire. Miles have been proven to be a valuable currency - there are endless possibilities when it comes to earning and redeeming. However, most miles expire and the airlines usually won’t notify the customer. MileageTracker manages miles across all airlines in one place. Our solution syncs all of the user’s miles and displays the balances with expiration dates in their online dashboard. The user is able to glance at all of their miles in one place, including information on when they expire. Custom alerts can be set to notify when miles expire by email.

Pitchbook enables client decision making by tapping into the power of data, available technology and market research, Pitchbook provides its clients with a News Feed section to provide current and consolidated news. However, the Company wanted to go a step further and provide customized news as opposed to several irrelevant articles that are surfaced in the News Feed. Our News recommendation engine allows us to affect more than 25,000 interactions a day on the Pitchbook platform by surfacing customized and relevant news articles using a combination of data science and topic modelling techniques

Interoperability, dispersed data, and inconsistent data formats are common issues across information science. This Open Data Literacy (ODL) Capstone tackles these issues in the realm of environmental restoration. Numerous restoration projects have been undertaken throughout the Puget Sound, but connecting investments in these projects to co-located indicators of habitat viability is challenging. Our team leveraged open, found data and open-source tools to build a scalable, sustainable data processing pipeline and an interactive, web-based visualization prototype. This helps our partners better tell the story of Puget Sound restoration efforts and demonstrates that open-source tools can help data curators meet open-data needs.

Rhema Life Ministry has an interesting information problem, in that they add a complex network of interpersonal interactions to the normal information challenges most organizations face. In my Capstone project, I built on their best practices by applying iSchool lessons about Information Grounds, Knowledge Management, Digital Curation, and of course Website Development. Software used included Photoshop, Lightroom, GitHub, VS Code and Nero. Communication skills and active listening were very important in allowing me to reach my goals, and exceed my sponsor’s expectations as we worked to develop a website that would support current success and future growth.

Computer Vision, an integral component in the development of Autonomous Cars, requires tens of thousands of tagged images for supervised learning. Currently, established and emerging companies must collect data from scratch, wasting time and money. They find that it is impractical to collate sufficiently geographically and semantically robust corpora effectively alone. Therefore, Road.io crowdsources dashcam footage from drivers which we then process and package in a Data-as-a-Service Model to Autonomous Car Companies. We seek to hasten the commercialization of Self-Driving Cars, become a premier resource for responsive, custom-filtered crowdsourced Data and empower everyone to have a global impact.

After the fall of the Ottoman Empire, Turkey experienced rapid language reformation that shifted written Turkish from an Arabic script to a Latin-based alphabet. Consequently, some texts were transcribed inaccurately or included subjective editorial decisions. These discrepancies are detrimental to the preservation of Ottoman history and culture. Scribe mitigates the faith gap between Arabic and Latin-based Turkish manuscripts by standardizing the transcription workflow. Our solution offers dynamic verification of transcriptions’ integrity by facilitating the comparison of original Ottoman text with Latin Turkish. Scribe maintains the authenticity of primary source texts in order to preserve Ottoman tradition and literature.

The UW has over 1500 hours of historically important audio in the Milo Ryan Phonoarchive. This collection includes archived CBS radio broadcasts, primarily from the World War 2 era. Currently, it is difficult for scholars to identify materials of interest within this collection. Metadata for this collection including titles, broadcast dates, and program titles were transcribed into a printed book. In this project, the book is being converted to a fully searchable digital finding aid using enhanced Optical Character Recognition (OCR) techniques to greatly enhance the original scans.