Der deutsche Education Blog

Microsoft Research Connections Blog

The Microsoft Research Connections blog shares stories of collaborations with computer scientists at academic and scientific institutions to advance technical innovations in computing, as well as related events, scholarships, and fellowships.

  • Microsoft Research Connections Blog

    Literome: extracting knowledge from biomedical publications

    • 0 Comments

    As any researcher knows, keeping up with scientific knowledge isn’t easy. This is especially true in the field of medical genetics, where advances in DNA sequencing technology have led to an exponential growth of genomics data. Such data hold the key to identifying disease genes and drug targets, because complex diseases inevitably stem from synergistic perturbations of pathways and other gene networks. Many of these interactions are known, but most of this knowledge resides in academic journals, the number of which has undergone its own exponential growth. It thus has become increasingly difficult for researchers to find relevant knowledge for genomic interpretation and to keep up with new genomics findings. Fortunately, help has arrived with the Literome Project.*

    Literome is an automatic curation system that both extracts genomic knowledge from PubMed (one of the world’s largest repositories of medical and life science journal articles) and makes this knowledge available in the cloud, with a website to facilitate browsing, searching, and reasoning. Currently, Literome focuses on the two types of knowledge most pertinent to genomic medicine: directed genic interactions, such as pathways, and genotype-phenotype associations. Users can search for interacting genes and the nature of the interactions, as well as for diseases and drugs associated with a given gene or single nucleotide polymorphism (SNP). Users can also search for indirect connections between two entities; for example, they can look to see if a gene and a disease might be linked by searching for known associations between an interacting gene and a related disease.

    Literome builds on Microsoft Research natural language processing (NLP) technology, extracting information from PubMed abstracts via our Statistical Parsing and Linguistics Analysis Toolkit (SPLAT), and uses the Microsoft Azure cloud platform to store, analyze, and disseminate the information.

    Scientists can use Literome in a number of ways, from exploratory browsing, to corroborating or refuting new discoveries, to programmatically integrating pathways and genotype-phenotype associations for making discoveries from genomics data. Literome is freely available for noncommercial use through an online service, or downloadable web services. It is our hope that Literome will help researchers search genomic medical findings that can lead to new understanding and treatment of genetically mediated diseases.

    Hoifung Poon, Researcher, Microsoft Research

    Learn more

    ____________________
    *The Literome Project is a joint project from Hoifung Poon, Chris Quirk, Charlie DeZiel, and David Heckerman of Microsoft Research.

  • Microsoft Research Connections Blog

    Internships pay big dividends

    • 0 Comments

    Each year, Microsoft Research Asia welcomes a new group of Korean interns, who spend three months or more conducting exciting research. Like all Microsoft Research interns, these talented young scientists derive several benefits from their internships, not the least of which is the mentorship of top Microsoft researchers in their field of interest, and exposure to the most advanced technology and best databases. Moreover, they get real-word experience as they collaborate with leading researchers on cutting-edge projects and innovative ventures. As the brief stories that follow reveal, internship is an exciting, mind-opening experience that reveals new opportunities for the interns.

    The challenges and achievements of internships
     
    Taesung LeeTaesung Lee, a PhD student at Pohang University of Science and Technology, has had three internships at Microsoft Research Asia, which have totaled more than a year and a half and have involved research in web-scale taxonomy cleansing and de-noising, and the attributes of knowledge bases.

    For Taesung, these internships have been invaluable experiences in the ups and downs of the research process. “I tried many ways to attack these challenges and explored positive and negative evidence,” Taesung recalls. “I learned that communication in research is extremely important; without it, the validity and feasibility of your project can be affected by your own prejudices.”
     
    While analyzing and solving problems, he said that he was very grateful for his collaborative partners’ help. When he encountered experimental failures, his mentors and colleagues gave him useful suggestions and support. All of which has helped him succeed, having two papers accepted for major conferences: VLDB 2011 and ICDE 2012.

    Taesung is especially appreciative of his Microsoft Research Asia mentor, Zhongyuan Wang. “Zhongyuan is not only my academic mentor but also my friend, always inspiring, giving me the courage and positive feedback to explore further. When we have different views on an issue, he offers many constructive suggestions. As I tread the road of research, Zhongyuan has given me a lot of positive energy.”

    Research with real-world impact
     
    Ji-Yong Shin“When people think of industry research labs in computer science, Microsoft Research is one of the first to come to mind. So when I first applied for an internship in 2008 while pursuing my master’s degree, Microsoft Research Asia was one of my top choices,” says Ji-Yong Shin, currently a PhD student at Cornell University.

    Like Taesung, Ji-Yong has served more than one Microsoft Research internship: in Beijing in 2008, Redmond in 2009, and Silicon Valley in 2011. Although it has been six years since I first interviewed Ji Yong for a Microsoft internship, he still remembers me asking him what kind of impact he thought he could make at Microsoft Research Asia.  

    “At that time, I was working on NAND flash memories, and I knew there were very few researchers at Microsoft Research Asia who were working in this area. So I said I could collaborate with the researchers and help them solve flash-related research questions,” Ji-Yong recalls. However, he learned the true meaning of “impact” after he actually started his internship.

    “I started working at the Platforms and Devices Center, which is currently part of the hardware computing group in Microsoft Research Asia, and I was surprised that the research topics my group were working on were not only academically meaningful but also produced results that other product or research groups wanted to use immediately. Before my internship, I was happy just writing papers in the lab and didn’t think much about how my research would affect people in the real world. Actually seeing others use my research results at Microsoft made me realize the real joy of research and made me think deeply about how to choose investigative topics that can make a difference in the world.”

    Learning on the rebound
     
    Youngho KimUniversity of Massachusetts Amherst PhD student Youngho Kim first interned at Microsoft Research Asia in 2007. In 2012, he obtained a second internship offer from Microsoft Research. When talking about his internship experiences at Microsoft Research Asia, Youngho describes it as “learning from rebound.” 

    “I was highly confident and proud of myself at first,” admits Youngho. “I was eager to prove my brilliance by successfully completing my project on semantic-level paraphrasing generation. However, halfway through the project, I realized that the data I was using, which consisted of web texts, had more noise than I could possibly filter out. I struggled to solve the problems by myself, but the data variation was too large to be generalized. Running out of time with no tangible results, I was devastated. Finally, I went to my mentor and confided my problem to him.”

    After much discussion, they agreed that Youngho should drop his own project and join his mentor’s for the remainder of his internship. “I was disappointed in myself. I had let my pride get the better of me. This experience forced me to take an honest look at my weaknesses and gave me a chance to grow further as a researcher.”
    Back at school, Youngho felt like a new man, as he enthusiastically initiated novel research, employing the new attitude he had acquired at Microsoft Research Asia.
    Youngho’s current internship project is going very well. “I understand that I have a long way to go before I’m a leading researcher in my area, but I’m a passionate scientist who’s always eager to learn. My Microsoft internship experiences will support me as I move ahead, and I appreciate the great opportunities they have given me.”

    The guidance of a top researcher

    Hyunson SeoWhen Hyunson Seo received email from Frank Soong from the Speech Group at Microsoft Research Asia, suggesting they talk about her internship application, she was thrilled. A PhD student in electrical engineering at Yonsei University, Hyunson still remembers her first phone call with Soong—who is a top researcher in Hyunson’s field—and her excitement at boarding a flight to Beijing for her internship in September 2011.

    “Though Microsoft Research interns enjoy many benefits and privileges, the most precious for me was the one-to-one mentoring system. I learned a lot from Frank. He always encouraged me and provided the big picture perspective that I might have missed. Whenever I encountered a problem, he was there to help me work through it. “

    “The project I took part in was about factoring speaker and environmental variability in speech recognition systems. I mastered many algorithms during the process and I learned how to achieve results in a short period.”

    “Knock on the door of Microsoft Research, and you will see the world,” Hyunson concludes.

    Long-term collaboration with Microsoft Research Asia
     
    Yohan ChonAs a recipient of a Microsoft Research Asia Fellowship, Yohan Chon was offered a four-month internship at Microsoft Research Asia, where he worked under Nic Lane and Feng Zhao in the Mobile and Sensing Systems (MASS) research group. That experience was the beginning of a long-term, ongoing collaboration with Microsoft Research.

    "I have had outstanding research experiences at Microsoft Research Asia, thanks to three great factors: my mentor, my team, and the environment. It’s really been an honor to get mentoring from pioneer researchers. They’ve inspired, guided, and encouraged me, making me a better researcher,” says Yohan. “It is also important to collaborate with passionate interns from other countries. Their desire and attitude have motivated me to work even harder.”

    Yohan continues to collaborate with researchers on the MASS team. “The internship at Microsoft Research Asia opened new opportunities for me. It has allowed me not only to improve my research ability but also to meet a great research team. I’m still working with researches in Microsoft Research Asia, and we’ve made great contributions for mobile sensing research communities.”

    Yohan’s final words capture the essence of the Microsoft Research intern experience: “It seemed a small opportunity at the beginning, but it significantly changed my research life.”

    Miran Lee, Principle Research Program Manager, Microsoft Research Asia

    Learn more

  • Microsoft Research Connections Blog

    Supporting the innovation pipeline

    • 0 Comments

    Through our collaborations with academic researchers around the world, Microsoft Research strives to harness the power and potential of computer science. We are constantly looking for new, creative uses of computing to help solve both theoretical and practical problems and engender economic and social benefits. This blog focuses on one such research project.

    Collaboration leads to quick and easy fabrication of circuit boards

    Yoshihiro Kawahara, an associate professor at the University of Tokyo, is devoted to solving real-world problems through the fusion of information science, hardware, and electromagnetics. Small wonder, then, that he received the 2014 Microsoft Research Japan New Faculty Award, which recognizes outstanding new faculty and encourages them to realize their potential in pursuing computer science research. Kawahara was honored for his work on building smart environments by using low-cost, intelligent sensors. To make such sensors more readily available, he has developed, in collaboration with Steve Hodges of Microsoft Research Cambridge, a technology that enables people to inexpensively prototype electrical circuits and sensors by using inkjet printers.

    Yoshihiro Kawahara displays his Japan New Faculty Award with Sadaoki Furui, president of Toyota Technological Institute at Chicago and chair of the award committee, on June 4, 2014, at Microsoft Japan.
    Professor Yoshihiro Kawahara (left) and Dr. Sadaoki Furui, president of Toyota Technological Institute at Chicago and chair of the award committee (right), display Kawahara’s Japan New Faculty Award plaque on June 4, 2014, at Microsoft Japan.

    Have you ever made your own electronic circuits? Unfortunately, the usual approach, which entails soldering components to a copper circuit board, poses a barrier for many beginners. Together with colleagues from Georgia Tech, Kawahara and Hodges proposed a much easier and more accessible technology by using a common inkjet printer.

    Their technology enables the creation of various electric circuits, touch sensors, and antennas by using common devices and materials. With the synergy of 3D printing and open-source hardware, this technology allows even inexperienced experimenters and hobbyists to make their own circuits quickly, easily, and inexpensively.

    A circuit that was printed on an inkjet printer by using the technique developed by Kawahara and HodgesThe circuit pictured on the left was printed on an inkjet printer, using the technique developed by Kawahara and Hodges. This innovative process is not only inexpensive, it also eliminates the tedious work of hand soldering the components.

    With the support of Microsoft Research’s CORE program and Masaaki Fukumoto of Microsoft Research, Kawahara and Hodges are expanding their work in hopes of creating fundamental tools and techniques to make double-sided and multilayer circuit boards, including mechanisms to fabricate through-hole vias, and they are implementing applications for .NET Gadgeteer.

    Turning research ideas into real business

    Beyond receiving accolades from the academic community, Kawahara and Hodges’ instant inkjet circuits are also opening up new business opportunities. Entrepreneur Shinya Shimizu was one of the first to see the technology’s potential to enable almost anyone to make their own circuit boards. He left his job at the consulting firm of McKinsey & Co., lined up an investor, and founded a startup called AgIC to develop and sell a toolkit based on the instant inkjet circuit technology. Shimizu demonstrated a prototype at the renowned South by Southwest (SXSW) festival in the United States last March, and he launched an experimental project in Kickstarter a month later, which received almost US$80,000 through crowdfunding. AgIC was selected as the Best Company as part of the Microsoft Innovation Award 2014, sponsored by Microsoft Ventures Tokyo, which will provide AgIC with further marketing and business alliance support.

    Shinya Shimizu, CEO of AgIC, and Shunichi Kajisa, CTO of Microsoft Japan, at presentation of AgIC’s Microsoft Innovation Award on May 30, 2014, in Tokyo
    Shinya Shimizu, CEO of AgIC (left), and Shunichi Kajisa, CTO of Microsoft Japan (right), at presentation of AgIC’s Microsoft Innovation Award on May 30, 2014, in Tokyo

    This work is just one example of how Microsoft is collaborating with talented individuals around the world on innovative academic research. We are very happy to see our academic collaborations resulting in practical applications that advance technology and will, we hope, help spark new ideas.

    Noboru Kuno, Research Program Manager, Microsoft Research Asia

    Learn more

Page 6 of 118 (353 items) «45678»