September 21, 2018

Microsoft Research (MSR) India: Call for Collaborative Projects on Cloud and AI Technologies for Societal Impact

Details on the suggested themes:

1) AI and NLP Technologies for Low Resource Languages
“Simply embedding endangered languages into the keyboards of smartphones will not save them. But, keeping these languages enmeshed in the fabric of daily life—which, particularly for the newer, younger speakers who are key to these languages’ survival, means being a viable way to communicate through technology—is the only way they will have even a slim hope of surviving.”, William Brenan, The New Yorker
Technology pervades all aspects of society and continues to change the way people access and share information, learn and educate, as well as provide and access services. Language is the main channel through which such transformational technology can be integrated into the socio-economic processes of a community.
However, this benefit is still limited to a subset of the world’s language communities and large populations worldwide are bereft of access to technology in their own languages. Most languages in the world lack the linguistic resources to build large data-driven (e.g., Deep Neural Net) models. To be able to truly support speech and language systems that can enable everyone on the planet, methodologies and techniques to build systems in resource constrained settings are essential.
With this goal in view, we invite project proposals in the areas of Natural Language Processing, Conversations and Speech that address the needs and aspirations of language-users currently unable to access technology.
We invite project proposals that seek to impact underserved communities through enabling language technology by creating economic opportunities, building technological skills, enhancing education and preserving local language and cultures for future generations.
Project proposals are invited in, but not limited to (1) Data: New/Innovative methodologies for data design and collection, e.g., gamification of data collection, crowdsourcing, etc, (2) Language Technology Systems: Designing new techniques and framework/architecture for technology for low resource languages, building Speech and NLP systems for low resource languages, and (3) Applications: At scale deployments of language technology applications that impact the community.
For more context, please see our Low Resource Speech Challenge (opens in new tab) at Interspeech 2018.

2) Ludic Design for Accessibility
“Man only plays when he is in the fullest sense of the word a human being, and he is only fully a human being when he plays” –Friedrich Schiller
We call our approach to the creation of technology solutions for accessibility as ludic design for accessibility. Ludic design holds that “non-productive” activities and desires—the need to play, to have fun, to be playful and the need for entertainment in one’s life— are central to the solution.
We are creating a Room of Empowerment at MSR India that will be a collaborative test bed focused on ludic design for accessibility. This Room will bring together persons with disabilities, researchers from industry and academia, practitioners, and NGOs together. We are inclusive in our definition of impairments to cover physical, sensory, neuro-muscular, and cognitive impairments across the spectrum.
We are seeking project proposals on creating solutions that enable people with disabilities to play and enjoy themselves with technology. Example ongoing projects in this area at MSR include video games for the vision impaired and gaze-controlled games and play for people with SSMI. Instead of narrowly focusing on specific tasks such as “enabling people with impairment X to do task Y”, we encourage projects that address a wide range of impairments and a wide range of tasks. The accepted project teams will be physically situated in the proposed Room of Empowerment. As mentioned, the vision for this Room is to harness technology for the sole purpose of empowering people with physical, neuro-muscular or cognitive impairments to have a roaring fun time playing.
Related projects at MSR include Avare (opens in new tab) (Augmented Virtual And Real Environments) and CodeTalk (opens in new tab) (Programming Tools for Visually Impaired Developers)

3) Inclusive employment, crowdsourcing, gig work

There is a significant fraction of the population that have low income or high debt. We believe that AI and cloud technologies have the potential to provide new employment and income opportunities to this population. We also believe that tools providing financial education, income and debt management and training has the potential to allow this underserved population to access basic financial services, such as bank loans.
Our goal is to improve the financial inclusion of low-income communities. Towards this end, we are inviting project proposals to investigate how technology can be used to 1) provide alternate sources of income (e.g., crowd-sourcing work), 2) educate low-income communities to better manage their finances,  3) provide skills that potentially allow them to break out of the cycle of poverty and 4) to improve gig work, for example, by promoting fair work conditions or building fair gig work platforms, or helping workers maximize their independence and flexibility, operate more competitively in the market, develop careers and so on.

Related projects at MSR include Karya (opens in new tab) (Dignified Digital Work) and our work in Digital Money, Financial Capability and Financial Inclusion (opens in new tab)

 4) Data science for societal challenges
Machine learning systems have been a success online. They help us find relevant information, discover interesting content and help us in our work and daily lives. As general tools for processing data, however, machine learning and data science can have broad applicability. What if we turned these techniques towards important societal challenges that we face today?
In many such problems, data is increasingly available. However, there are unique challenges on establishing data provenance, removing historical biases, and distinguishing correlations from causal factors. We invite project proposals for creative use of machine learning and data science to tackle societal problems, including, but not limited to applications in education, governance, financial inclusion and sustainability. We especially encourage inter-disciplinary projects with collaborations between technologists and domain experts.Example of a related project at MSR: Early Warning systems for preventing school dropouts

5) Technologies for Health and Well-being
In many parts of the world, people struggle with limited access to healthcare. According to a report by The World Health Organization, at least half of the world’s population is unable to receive essential health services. Doctors and healthcare providers face their own challenges as they deal with an ever-increasing number of patients with chronic ailments, new communicable diseases and mental health issues.
Help us reimagine the future of health and well-being through innovative uses of technology. We invite project proposals for all facets of healthcare, including community-based and institutional healthcare, and at all stages including diagnosis, treatment and prognosis. We specifically encourage projects which combine technological ideas and social engineering, and look at these issues from multiple perspectives to solve these problems.
Relevant themes include, but are not limited to medical decision-making, adherence and monitoring outcomes, mental health and public health awareness
Example of a related project at MSR: Tracking adherence to tuberculosis medication through 99DOTS (opens in new tab)

6) Public Awareness and Combating Misinformation
“Right information at the right time”
Access to information is a powerful force for alleviating societal inequality. Today, however, we face big challenges in realizing this goal: on the one hand, public agencies face difficulties in spreading vital information on health, education and livelihood; on the other, misinformation or “fake” news is travelling faster than before and leading to grave outcomes. How do we channel modern tools of communication to promote awareness on critical issues and reduce misinformation?
We invite project proposals that intersect the themes of information access, public awareness, and misinformation for societally-critical domains such as health, education, livelihood, news and policy. Topics include, but are not limited to:
• Enhancing the reach of public awareness programs
• Reimagining public awareness through ICT and social media technologies
• Combating misinformation in social media and identifying veracity of information
• Methods for establishing trust in provided information
Example of a related project at MSR: Learn2Earn, which is a social incentives-driven mass awareness system

7) Other systems and tools for societal impact
Under this category we invite project proposals that look at the design and development of systems, infrastructure and tools (using cloud and AI technologies) for societal impact, yet not explicitly covered by the previous themes. The projects would ideally be domain-neutral, or begin with a specific focus, but have components that would be generalizable across domains. Example areas include overcoming low-literacy, rural networking, building blockchain, improving sustainability, etc.
Example of related projects at MSR include VideoKheti (opens in new tab) (making video content accessible to low-literate farmers) and Vishrambh (opens in new tab) (Trusted Philanthropy with end-to-end transparency).