research

The Interface Ecology Lab fosters integrative research projects spanning hardware, software, and theory, producing natural user interfaces, creativity support environments, games, interaction techniques, visualization algorithms, semantics, programming languages, interactive installations, and evaluation methodologies.
Embodied interfaces research integrates embedded systems with HCI and the interface ecosystems approach, to imagine and develop new embodied experiences through computing and signal processing. We create new devices and experiences, using low power sensors, computers, and networking. Context is modeled based on recognized patterns. Affordances are designed; systems responds with clear feedback and media, to stimulate particpants' senses.
Information composition is a holistic, visual, and semantic medium for curation of digital objects, supporting information-based ideation. Composition is non-linear and freeform, breaking out of lists and grids to support the synthesis of diverse ideas and emergence of new ones. Composition integrates rich media curation with semantic metadata and expressive annotation through text and sketching.
Information-based ideation encompasses human activities in which generating new ideas is essential, while using information as support. IBI spans imagining and planning an evening, vacation, makeover, summer internship, career, thesis, invention, service, company, crisis response, and approaches to interdisciplinary topics such as sustainability. We show how media of curation support IBI. Evaluation integrates qualitative methods, such as grounded theory, with derivation and application of scalable quantitative ideation metrics of curation.
Trans-surface interaction addresses moving information objects across multi-display environments that support sensory interaction modalities such as touch, pen, and free-air. It is characterized by user experiences of continuously manipulating information across interactive surfaces, providing a sense of connection.
BigSemantics is an open source software architecture for developing powerful applications that present interactive metadata semantics from diverse information sources. BigSemantics provides a meta-metadata language for defining metadata types, a type system for reusing and extending new types, a repository of types that involves many useful sources and use cases, runtime libraries for conveniently accessing and using metadata semantics in applications, and a web service that facilitates thin clients. Based on BigSemantics, we developed Metadata In-Context Expander (MICE), an example web application that enables exploring linked semantics in the current context.
TweetBubble is a chrome extension that helps twitter users follow associational chains of tweets through #hashtags and @users. People can experience and relate a variety of content in-context helping them develop multiple perspectives on a topic. TweetBubble makes browsing a more fun and fluid experience.
TweetBubble uses BigSemantics and Metadata In-Context Expander.



prior projects

S.IM.PL is our open source initiative for augmenting popular programming languages to facilitate inter-operation, with an emphasis on simplifying development of distributed applications that represent the world. The foundation layer, S.IM.PL Serialization, is based in a language-independent type system encapsulated by TranslationScopes, and enables cross-language code generation, as well as de/serialization to XML, JSON, and TLV. Object-Oriented Distributed Semantic Services (OODSS) simplifies building distributed applications and services, which decode message passing using S.IM.PL Translation Scopes, to connect iPhone, Android, Java, and .Net clients to Java and .Net servers. BigSemantics also uses S.IM.PL.
We are developing new digital games for Teaching Team Coordination. This research takes ethnography of fire emergency response work practice as a basis, and develops simulations of aspects of team coordination. Real-world actions map to game activity, as players work together to find real goals and avoid virtual hazards. We invented Zero-Fidelity Simulation, a method that focuses on reproducing the communication and information distribution components of target environments to produce engaging, low-cost, and effective educatioal experiences.
Hurricane Recovery: Collecting Locative Media to Rebuild Local Knowledge Engages in an iterative participatory process of reaching out to evacuee communities subsequent to Hurricane Katrina, gathering information about needs and desires, building situated semantics and a locative media collection sensemaking system, and using the system to collect, organize, and re-present images, interviews, and metadata. Digital photographs are connected with GPS sensor data, semantics, a zoomable map interface, and an image clustering algorithm.
Test Collection consists of a set of documents, a clearly formed problem that an algorithm is supposed to provide solutions to, and the answers that the algorithm should produce when executed on the documents. The present research develops an open source Test Collection Digital Library System. The system enables collecting and labeling documents, and publishing the resulting test collections.