Collaborative Programs

Collaborative Programs

Risk Factors for stroke in an Indian population and prediction algorithm for first-ever stroke outcome

The project with Manipal University entitled “Risk factors for stroke in an Indian population and prediction algorithms for first-ever stroke and stroke outcome” is trying to determine the risk factors for stroke in an Indian population and the prediction algorithms for first-ever stroke and stroke outcome. As many may be aware, strokes are perhaps the second leading cause of death and a major cause of neurological disability in the world. The total cost of stroke to USA alone is estimated at $71 billion per year (in 2012). Stroke is not limited to western or high-income countries only: about 85% of all stroke deaths are registered in low- and middle- income countries. Improved prevention and management of stroke can have tremendous impact on both economy and health. Nearly 70% of the strokes are first-ever strokes and so, prevention of stroke can considerably reduce morbidity and the financial burden of patients and healthcare institutions. Stroke impairs many critical neurological functions, causing a large number and broad range of physical and social disabilities. Due to the complex nature of stroke recovery, clinical outcomes after stroke can vary considerably - from complete recovery to permanent disability and death. To tailor an effective treatment for an individual, it is important to assess the risk of damage and chances of potential recovery. Stroke outcome prediction can be used to guide treatment decisions for a favourable clinical outcome.

The goals of this collaborative research project broadly are to:

  • Study risk factors of stroke for the Indian population
  • Study new genetic and blood markers for stroke
  • Design prediction algorithms for first-ever stroke and stroke outcome Findings from the project could potentially be used for designing tools for preventive medical care, building decision systems for choosing treatment options and developing technologies for personalized healthcare management

Context-Aware Large-Scale Mobile Crowd-Sourcing

The project with Singapore Management University entitled “Context-Aware Large-Scale Mobile Crowd-Sourcing” draws on two active disciplines of computing science research, namely mobile sensing/analytics and decision optimization to address aspects such as context sensing, context prediction and utility maximization of the worker as well as that of the mobile crowdsourcing platform.

Some background is provided here before the project is described in some detail. Commercial mobile crowd-sourcing solutions (such as TxtEagle and Ushahidi currently focus on adapting ‘online’ tasks (such as OCR) to device-specific characteristics (e.g., screen size). For mobile crowdsourcing to be truly ubiquitous, such a service needs two logically orthogonal attributes:

  • It should allow users to opportunistically complete conventional tasks using their mobile devices as part of their daily lifestyle-driven activities.
  • It should leverage upon natural movement trajectories/context of individuals to support an emerging category of spatiotemporally-constrained tasks tied to specific real-world locations. e.g., verifying the queues at a movie theatre in next 30 minutes, ascertaining the inventory of a product at designated stores, taking pictures of real-estate locations.

For such mobile crowd-sourcing paradigms, the current and future context of the crowd worker becomes a critical determinant of both the a) worker’s task selection/assignment process and b) the efficacy of the task execution. Recent advances in mobile sensing and sensor data analytics now enable us to dynamically sense and infer a wide variety of such dynamic individual context (e.g., if the person is walking or standing, or if she is in a noisy vs. quiet environment). Moreover, predictive analytics applied on such rich context data also enables such individual context to be predicted at future instances (e.g., if the person is likely to encounter more than five minutes of wait time in the taxi queue or the likely path she will be taking after work on a Thursday).

The proposed multi-year research effort addresses both mobile sensing/analytics and decision optimization questions from a theoretical perspective and then validates the proposed solutions empirically. For the inference and prediction of mobile context, we shall investigate advances in both real-time stream analytics and large-scale data analytics algorithms, applied over sensor streams generated by mobile devices. A key challenge will be to preserve the accuracy of the inference process, while minimizing the potentially prohibitive energy overheads of sensing on mobile devices. Our proposed work will not only result in theoretical advances in these areas, but will also empirically evaluate and demonstrate their usefulness over a relatively large set of mobile urban participants. Such large-scale experimental validation will be enabled by leveraging on the globally-unique LiveLabs Urban Lifestyle Experimentation Platform at SMU, which will provide near-real time, deep context for a large set of participants at multiple urban public spaces in Singapore

A Complex Systems Approach for Predictable Job Completion in Business Process Crowdsourcing

The project with the Indian Institute of Science entitled “A Complex Systems Approach for Predictable Job Completion in Business Process Crowdsourcing” tries to develop theoretical models for crowd-sourced tasks for guaranteed completions.

To give some background, crowdsourcing is defined as the act of outsourcing tasks, traditionally performed by an employee or contractor, to an undefined, large group of people or community (the “crowd”) through an open call. Crowdsourcing, typically, is best suited for tasks that require human skills and intelligence, and has been successfully applied to a wide variety of tasks such as content generation, software development, creative design, market research and R&D problem solving. When it comes to crowdsourcing human tasks related to an enterprise, it poses several challenges like security, confidentiality, performance, and timely completion.

The focus of this research proposal is the problem of predictable job completion in crowdsourcing. In this project, we propose to develop theoretical model for predicting completion time of a task taking into account for crowd behaviour and how optimally we can schedule tasks to ensure guarantees about completion times.


Faculty Sabbaticals

XRCI welcomes faculties from universities across the globe working in the areas of mutual interest to pursue sabbatical with us. It gives them the opportunity to work closely with our researchers, and have the opportunity to explore collaborative research projects projects at industrial scale, work with state-of-the-art technology, and experience Xerox. Researchers at XRCI are focussing to capture innovation opportunities in emerging markets and to advance Xerox’s position as the leading global provider of document and business process services. Four of the key areas of focus at XRCI include Education, Healthcare, Smart Cities and Banking. This engagement benefits both professors and the Xerox researchers as new and exciting ideas are discussed and shared. It is often the case that faculty return to their universities with new research directions and educational ideas.
These sabbatical opportunities are available throughout the year, we typically offer the positions for three months to one year.
We are currently accepting faculty applications by invitation only. You should reach out to your Xerox collaborator to apply for a sabbatical position at XRCI.


Seminar Talks

Seminar talks happen to be a useful way to exchange research ideas and XRCI promotes a two way communication where researchers from academic institutes are invited to give lectures and talks while researchers from XRCI are encouraged to do the same in academic institutes.