Launched in 2006, the IBM Open Collaboration Research (OCR) awards program provides a new model for partnerships between universities and IBM to study significantly challenging research topics impacting the 21st century.
IBM Research and Carnegie Mellon University will develop new algorithms to discover socially salient points-of-interest and build models of physical environment using first-person vision and massive crowd sourcing and implement such algorithms as reusable open source components by foreseeing applications such as navigation of persons with disabilities and senior citizens, and big event management.
IBM Research and Rice University will develop free, open-source, question generation algorithms that take educational texts as input and produce high quality multiple choice questions as output, and to empirically determine whether software-generated questions approach the usefulness of subject-matter-expert generated questions. Such algorithms would be beneficial for a wide variety of uses within educational technology, and would help to greatly reduce the costs required to deliver high quality practice opportunities to students. A critical first step is open-source software to determine to what extent software-generated questions work as effectively as subject-matter-expert-authored questions in reinforcing what should have been learned and assessing full comprehension.
IBM Research and Boston University will develop software discovery and change detection technologies for cloud. We will apply these discovery techniques for security analysis, compliance and deployment optimization use cases. We will develop Configuration Analytics and Malware Signature Analysis solutions for Cloud Operational Analytics.