Microsoft Research is pleased to announce an invitation for proposals for collaborative research (CORE) projects, with awardees receiving funding for the 2013 fiscal year. Microsoft Research sponsors collaborative research projects in Japan for local researchers to advance the state of the art in research, inspire technological innovation, foster young researchers, and to establish Microsoft as a valuable research and technology partner for higher education.
CORE Proposal Submission Process
On This Page
1. Applicants
The following are eligible to apply for this award:
- Faculty or researchers from a Japanese University that awards degrees at the baccalaureate level or above by MEXT
- PhD holders or PhD candidates in related research areas who are currently working at a university or institute in Japan
- Junior faculty and young researchers who obtained a PhD within the last five years and are currently working at a university or institute in Japan are strongly encouraged to apply
- New applicants who have not received a CORE award during the past year
Notes
- An applicant may not submit more than one application.
- Award recipients from the prior year are not eligible for this award.
2. Research Topics
The proposed research should relate to current Microsoft "Research Areas" listed on the following pages:
Additional Topics of Interest
We welcome proposals that demonstrate innovative research and application—particularly in the following areas.
Mobile Sensing
(Microsoft Research contacts: Jacky Shen, Nic Lane)From the perspective of interdisciplinary research, we encourage proposals related, but not limited, to the following topics:
- Mobile sensory data back-end processing: large-scale sensory data that is collected from mobile sensor networks in the cloud, which includes data storing, sharing, querying, and visualization. By taking advantage of the cloud computing infrastructure, it is possible to integrate large-scale sensor networks with sensing applications to perform pervasive computations.
- Mobile phone-based application design: mobile phone applications that are built on top of Microsoft Research's SenseWeb. All SensorApps need to be shared on SenseWeb to be available to the public as free downloads.
- Participatory sensing: the use of massive network-accessible mobile devices (for example, smartphones) acting as mobile sensor nodes with the capability to collect and share data. These mobile devices transfer sensory data to the cloud for storage and processing by using interactive applications and tools to query, analyze, or visualize data.
The following are a few examples of mobile sensors or sensor nodes that are mentioned in the above section:
- Smartphones: act as mobile sensors that have capabilities such as taking photos, recording sounds, and sending geo-location information based on GPS, Wi-Fi, cellular towers, and so forth. New sensing functions can be appended to the smartphone easily to obtain additional data.
- Body sensors: carry-on sensors that can sample and transmit data—such as heart rates and body temperature—that are associated with a person’s activities.
- Vehicular sensors: sensors on vehicles that communicate with another vehicle or with a roadside object, thereby enabling intelligent transportation controlling and monitoring.
Available Microsoft technologies:
- Platforms and operating systems: SenseWeb, SensorMap, Windows Mobile, Windows Phone, Windows Embedded, Windows operating system, Windows Azure
- Development tools: Microsoft Visual Studio, Microsoft Visual C++, Microsoft SQL Server, Windows Server, Microsoft Silverlight
Linking Language with Knowledge and Context
(Microsoft Research contact: Junichi Tsujii)Computer systems have been increasingly recognized as intelligent agents that participate in creative endeavors with humans or help humans engage in such endeavors. Computational processes that link language with knowledge and context are crucial to enhance computer systems’ ability to act as intelligent partners.
- Cognitive modeling of language understanding: the process of understanding languages is a typical process of information integration. The meanings of language expressions—in addition to other modalities such as gestures and audio-visual signals—are integrated to under sentences fully, in either speech or text. Eye-tracking experiments have revealed that very early stages of language processing, such as parsing, are affected by contextual information. The ability to model computational processes in human language processing is the key to enable the development of more flexible, natural, and dynamic interaction systems in the future.
- Computing intra- and inter-textual structures: sentences are basic units of propositions, and the concept of how structures and meanings of sentences can be computed has been studied intensively in computer science and linguistics. However, there is very little research on how to capture structures and meanings beyond the sentence level. Such research would be essential for the intelligent summarization of multiple texts, chronological development of ideas in a community, and systematic organization of information in a set of related texts, and so forth.
- Knowledge discovery and hypothesis generation from text: one of the greatest challenges in text mining is to develop a system that can discover and reveal new pieces of knowledge that a set of given documents contains. The application of such data mining techniques to text has not been successful to date. The combination of information extraction techniques, such as named entity recognition (NER) and relation extraction (RR), to map text to existing knowledge by augmenting or changing existing knowledge is increasingly important. Related research includes “understanding by reading,” “contradiction discovery,” and “textual entailment.”
- Intention recognition from limited input: a major challenge in information retrieval, question-answering, and dialogue systems is how to identify users’ intention in a short utterance and/or a small set of keywords. One has to enrich users’ input and identify users’ implicit intention by considering the context of the users. The context includes preceding utterances, domain knowledge, and user profiles.
Probabilistic Knowledgebase and Its Applications
(Microsoft Research contact: Haixun Wang)“Big data” is no longer a buzz word or a catch phrase. It’s a reality. To leverage big data, we must have a better understanding of its content and structure.
- Knowledge-empowered text analytics: at Microsoft Research, we have built a knowledge base or semantic network known as Probase. Compared with other knowledge bases, Probase is unique in two aspects. First, Probase has an extremely large concept consisting of approximately 2.7 million categories of data and has the potential to support a wide range of applications. For instance, it will greatly enhance the performance of text mining and natural language processing, as a machine with common knowledge of humans to better understand communications. We encourage proposals on related topics of information extraction, knowledge base construction, knowledge inferencing, knowledge-based natural language processing, text analytics, and other related areas.
- Knowledge-empowered text analytics: Trinity is an in-memory, distributed graph platform developed at Microsoft Research. At the heart of Trinity is an in-memory key-value store. As an all-in-memory key-value store, Trinity provides fast random data access. This feature naturally makes Trinity suitable for large graph processing. Trinity is also a graph database from the perspective of data management. It is a parallel graph computation platform from the perspective of graph analytics. As a database, it provides features such as data indexing, concurrent query processing, and concurrency control. As a computation platform, it provides vertex-based parallel graph computation on large-scale graphs. Trinity enables a variety of applications that need to deal with graph data. We look forward to proposals on the topics of graph management and mining.
Machine Learning and Game Theory
(Microsoft Research contact: Tie-Yan Liu)At Microsoft Research, we are advancing the state of the art in research on Internet economics and computational advertising through a combination of basic and applied research. We are also collaborating with the world’s top researchers and institutions to shape the future of this research area.
We are looking for research proposals related, but not limited, to the following topics:
- Machine learning
- Online learning
- Reinforcement learning
- Adversarial learning
- Game theory
- Mechanism design for new Internet applications (for example, sponsored search, cloud computing, and application marketplace)
- Behavioral/evolutional game theory
- Equilibrium analysis and equilibrium finding
Exploratory Search
(Microsoft Research contact: Tetsuya Sakai)For many information needs, the one-query, 10-blue-links model is not sufficient. Information needs may be vague and may change. For some clear queries, systems may be able to provide the direct answers but for others, human-computer interaction for clarifying and satisfying the need may be inevitable. Exploratory search refers to the interactive and seamless nature of query formulation and search result presentation in the following:
- Interaction in diversified and aggregated search: diversified search tries to accommodate different user needs for ambiguous and underspecified queries. Aggregated search combines search results from different information sources and media. But if the system understands the user better through interaction, it may be able to improve its presentation strategies.
- Interaction in search result summarization and visualization: at times, just collecting and presenting different search results is not sufficient. The system may have to summarize the results or even visualize them to fit the display size or to satisfy the user very quickly. What are the effective ways to summarize? When is visualization effective? What is the role of human-computer interaction?
- Assisting query formulation and reformulation: the user's information needs may be vague and difficult to express. The user's physical environment may make it difficult to input queries. The user may choose to specialize or generalize queries or may change to another topic. How can the system assist the user in such cases?
- Anticipating information needs: human-computer interaction for search is not necessarily triggered by the user. Systems can proactively start helping users. In order to do this, the system needs to watch the user's context and anticipate their needs. Can we build a non-intrusive system that knows exactly when and how to help the user?
Mobile + Local + Social
(Microsoft Research contact: Xing Xie)With the rapid evolution of social media, location-based services, sensing technologies, and mobile devices, the integration of mobile + local + social (SoLoMo) starts to enable numerous innovative services on the Internet. With the great success of Facebook, Twitter, Foursquare, and many other public services, the research community is also thinking about the future of SoLoMo.
- Location-based social networks (LBSN): social networks have been prevalent on the Internet and have become a hot research topic, attracting many professionals from a variety of fields. By adding a location dimension, we can bring online social networks back to the physical world and share our real-life experiences in the virtual world conveniently. In LBSN, people can not only track and share location-related information with each other via mobile devices and desktop computers, but they can also take advantage of collaborative social knowledge acquired from user-generated location-related content. Using a user’s location, LBSN bridges the gap between online societies and the physical world and enables many novel applications that can change the way we live, such as in travel planning, location/friend recommendations, community discovery, human mobility modeling, and user activity analysis.
- Urban computing: an emerging concept, urban computing promotes the use of every sensor, device, person, vehicle, building, and street in urban areas as components to enable city-wide computing services for local residents and businesses. Urban computing aims to enhance both human life and the urban environment in an intelligent manner through a recurrent process of sensing, mining, understanding, and improving. Urban computing also aims to deeply understand the nature of, and science behind, the phenomena occurring in urban spaces, by using a variety of heterogeneous data, such as traffic flow, human mobility, geographic and map data, environment, energy consumption, population, and economics.
- Managing big data from the physical world: context awareness is a key concept in ubiquitous computing. Computing systems become more intelligent through analyzing and reacting to the physical world surrounding them. By accumulating and aggregating large-scale, physical world, contextual data from multiple users and multiple devices over a long period, we can obtain collective social intelligence from them. Based on this information, more innovative Internet-based services can be developed to facilitate people’s everyday lives. We can link data generated by different people, services, or sensors together, based on a unified knowledge model, and provide the intelligence as a service in the cloud.
Human-Computer Interaction
(Microsoft Research contacts: Hong Tan, Xiang Cao, Koji Yatani)The Human-Computer Interaction group at Microsoft Research invites researchers to collaborate with us in creating the next generation of natural user interfaces. We encourage proposal submissions on, but not limited to, the following topics:
- Novel input/output modalities and hardware: novel hardware technologies can facilitate the use of computer devices in various contexts and enhance the user experience. This area investigates novel input and output technologies for interactive systems. In particular, it focuses on novel hardware to extend the current input/output capabilities of computer devices. Example projects are as follows:
- Interactive devices and systems using haptic feedback
- Non-visual/audio output modalities (for example, temperature, olfactory)
- Hardware to support novel input capabilities
- Sensing technologies and applications for ubiquitous and wearable computing: ubiquitous computing poses many interesting technical challenges to make the computer technology pervasive. Sensing users, activities, and contexts have not matured yet. New applications for ubiquitous computing technologies are not fully investigated. This area explores new sensing technologies for ubiquitous and wearable computing and builds applications that use these technologies. Example projects are as follows:
- Novel wearable sensors for ubiquitous computing environments
- Gesture/activity recognition using sensors
- Systems for understanding user’s high-level activities through sensor fusion
- Applications using sensors in various user contexts and locations (from indoor to outdoor, from workplace to home)
- Visual analytic tools and systems: as data become diverse and huge, interactive systems now have a more important role in helping the user explore the data, extract important information, and perform sense-making or decision-making. In addition to data, analysis based on theories can also enhance the understanding of the phenomena the user is facing. This area investigates interactive systems to support the user’s sense-making and decision-making through data and theories. Example projects are as follows:
- Visualization of large-scale data for sense-making activities
- Data-driven and/or theory-driven analytic tools
- Decision-making support on mobile devices
- Mobile interactions and applications: many users now enjoy the power of mobile computing. The user experience could be enriched by improving mobile device interfaces and developing new applications to use on mobile devices. This area explores new ways to interact with mobile devices and their applications. Example projects are as follows:
- Novel interface designs for mobile phones
- Mobile applications enhanced by user’s contexts, social network, and sensor data
- Remote collaboration and communication systems: computer technologies can enable, facilitate, and greatly enrich collaboration and communication among people who cannot be in the same location. This area aims to investigate interactive systems to help people communicate and collaborate remotely. Example projects are as follows:
- Interactive systems enhancing communications among family and friends
- Remote collaboration systems using tangible interfaces, multi-modal systems, and/or interactive robots
- Systems for supporting creativity and information organization: people often need information and inspiration to accomplish tasks that require creativity. Easy access to necessary data could help the user’s thinking process. This area explores tools and systems to support the user’s creativity and information organization. Example projects are as follows:
- Systems for supporting creativity-related activities (for example, storytelling, brainstorming, information gathering)
- Information visualization tools for various types of media and data
3. Microsoft Researcher
In your proposal, state the Microsoft Research researcher with whom you propose to collaborate, as well as how the collaboration will occur (for example, meeting, workshop, short stay at the Microsoft Research Asia lab, or dispatch of interns).
4. Research Grant
The grant will be awarded as funding to the university or institute on behalf of the awardee for their selected research project. Unless otherwise arranged for outcomes, a memorandum of understanding clearly stating the following four points will apply. Please specify in the proposal a special arrangement for outcomes, if any. The receiving party of the funding must agree to the following requirements:
- The award shall be used solely to support basic research in the area of research detailed above; and
- All results derived from the research must, as soon as they are generated, be readily put in the public domain, freely and without restrictions, and accordingly the University shall waive all proprietary right, title, and interest in and to such results
- This grant must be acknowledged in all publications, press releases, and other publicity that is connected with the research program detailed above.
- The results must be presented to the IJARC Academic Advisory Committee and Microsoft Research at an annual CORE project review meeting, which tentatively takes place in 2013 at a date to be announced. The selected winners are also highly encouraged to attend the CORE review meeting of the prior year's projects, which will be held in June 2012.
5. Duration and Size of Project
- Duration/starting date: Basically one year from November 2012
- Funding amount: From JPY 1,500,000 to JPY 3,000,000, per project
- Number of awards: Based on the quality of the proposals, approximately 10 to 15 projects will be selected
6. Proposal Format
Proposals should be written in English with full details of the project. The proposal should be seven pages maximum, typed in 10 pt. font or larger, double-spaced, and in either Microsoft Office Word or PDF format. Proposals that do not meet these requirements will be automatically excluded from consideration.
Provide the following information about your proposed research:
- Problem Statement: What is the problem area addressed by the proposal and why is it important? What is the potential contribution to the field of the project if successful?
- Expected Outcomes: What tangible assets, if any, will be created or produced as a result of the proposed project?
- Research Schedule: When is the project to be completed? What milestones will be used to measure progress of the project and when will they be completed?
- Collaboration Plan: If any collaborative opportunity with Microsoft Research Asia researchers is desired, clearly state specific collaboration methods (for example, meeting or workshop, short stay in lab, or dispatch of interns).
- Budget Planning: Provide a budget breakdown that describes how the award will be used, including hardware or software purchases, salaries, and other costs.
- Use of Microsoft Technologies: Describe the Microsoft tools and technologies (if any) to be used in this project. While the use of Microsoft technologies is not a condition of this invitation for proposals, any proposal relying exclusively on non-Microsoft technologies should provide a justification for why this must be the case. We encourage the use of Windows Azure for the deployment of cloud applications. Use the Windows Azure pricing calculator to estimate the resources you need, and account for for your results in your budget breakdown.
- Related Research: Give a brief summary of the current state of the art in this field, including references where appropriate.
- Qualifications of Principal Investigator: Include a brief description of any relevant prior research, teaching, publication or other professional experience. A detailed vita or list of publications is not required.
7. Application Form
Complete the application form and submit it by the application deadline of 15:00 (3:00 P.M.), September 27, 2012.
8. Selection of Awardees
- First selection – The selection will be made by the Academic Advisory Committee and Microsoft Research. Results will be provided by email. The selected researchers are expected to attend the second selection interview as follows.
- Second selection – Interviews at the Microsoft Japan office in Tokyo in November or December 2012. The interview consists of a 15-minute PowerPoint presentation and 10 minutes of questions and answers. The Academic Advisory Committee and Microsoft Research will interview the applicants and determine the finalists. The official announcement will be made in January 2013.
9. Personal Information
Please read "Handling of personal information" in the cover page of the proposal before submission.
10. Application Deadline
The submission deadline is 15:00 (3:00 P.M.), September 27, 2012.
- Japan CORE Project Workshop 2012
Tokyo, Japan ·29 June 2012 - Japan CORE Project Workshop 2011
Beijing, China ·31 May 2011 - Japan-Taiwan Academic Forum 2010
Taipei, Taiwan ·17 May 2010
