European Commission High-Level Expert Group on AI
Fredrik Heintz has been appointed to the new European Commission High-Level
Expert Group on AI (AI HLG).
The AI HLG will have as ageneral objective to support the implementation of the European strategy on AI. This will include the elaboration of recommendations on future
AI-related policy development and on ethical, legal and societal issues related to AI, including socio-economic challenges.Moreover, the AI HLG will serve as the steering group for the European AI Alliance's work, interact with other initiatives, help stimulate a multi-stakeholder dialogue, gather participants' views and reflect them in its analysis and reports.
This is a great opportunity to take an active part in the European AI efforts and to influence the work. It is also another indication of Linköping University's leading role in AI.
Distinguished teaching achievements 2018
Marco Kuhlmann, Associate Professor of Computer Science at IDA, received the 2018 award for distinguished teaching achievements from the Institute of Technology. The award acknowledges Marco's work with the course Mathematics on the Bachelor's programme in Innovative Programming as well as the development of several new Bachelor's and Master's level courses in natural language processing.
More efficient bug handling in large-scale software development
In a recent PhD thesis Leif Jonsson has studied how one can use machine
learning (ML) to make bug handling in large-scale software development more
efficient. This was done by empirically studying which ML techniques are
suitable to use and which accuracy is reachable, but also implementing a
mathematically correct and very efficient version of a very popular ML
technique for Topic Modeling. Based on this, a new Bayesian classifier called
DOLDA is developed.
Topic Model Inference for Textual Data
Textual data, such as news items, literature, and political speeches are today
available in digital formats. Probabilistic topic models is a versatile class
of models to analyze topic compositions in corpora. Måns Magnusson has in his
thesis developed scalable and efficient methods to enable statistically
correct, large-scale inference for very large corpora.
Marco Kuhlmann gets the Distinguished Teaching Award
Each year since 1990, the Union of Technology and Science Students awards the "Golden Carrot" for distinguished teaching on the programmes at the Institute of Technology. This year, the winner of this distinction is Marco Kuhlmann, Associate Professor of Computer Science at IDA, who had been nominated by the students in Computer Science and Engineering.
Prize for the best Bachelor and Master theses 2017
The computer science department together with the computer society in Sweden
has for the 19th time awarded the annual prize for the best Master and Bachelor
theses in 2017. The winners were among 6 thesis that were nominated this year.
Alexander Ernfridsson was awarded for the Bachelor level and Tova Linder
together with Ola Jigin at the Master level.
Best thesis prize awarded by ITSM
Two Information Technology students, Rasmus Lindström and Nicklas Östman, were
awarded the best Master thesis prize for their work on how agile development
scales in different types of organisations. The prize was awarded by ITSM, the
knowledge network for active professionals within IT service management.
The thesis studies how the agile software development process scales in different types of organisations including those with a rigid and hierarchical steering when defining requirements. The prize was awarded due to the work "being of a high scientific quality and a strong grounding in both scientific literature and empirical evidence."
Programming and Optimization Techniques for GPU-based Systems
Today's computer systems are increasingly heterogeneous, where one or several
general-purpose processors (CPU) are complemented by hardware accelerators such
as graphics processors (GPU) that can perform certain computations faster
and/or more energy-efficiently. But writing and optimizing portable programs
for such heterogeneous systems is notoriously difficult.
In his PhD thesis, Dr. Lu Li has developed a number of techniques and software tools that can help the programmer with this task. He has proposed a number of techniques and tools to help the programmer in:
- writing portable programs for heterogeneous computer systems,
- automatically optimizing the program execution flow as well as data transfers and memory management,
- modeling optimization-relevant platform properties for program adaptivity and tool retargetability, and
- conveniently and portably measuring time and energy consumption of program parts in heterogeneous systems.
Honoring ceremony for the Head of the Department
IDA's departing head of the department, Professor Mariam Kamkar, was honored at
a ceremony held during the institution's Christmas dinner. Mariam took office
in 2000, and by the end of 2017/2018 she will have been prefect for 18 years,
more than half the time that IDA has existed. In 2008, she was the inaugural
recipient of the University Leadership Prize.
The farewell ceremony included speeches by Sweden’s first computer science professor, Erik Sandewall, LiTH's Dean, Ulf Nilsson, and incoming prefect, Henrik Eriksson. The speakers highlighted Mariam's wise leadership, her commitment to gender equality, and the enthusiasm she brought to the role of department head. They spoke warmly of her forthright interpersonal style, her conflict resolution skills, and her emphasis on the importance of all new international employees learning Swedish. Gifts from IDA staff were presented during the ceremony, along with thanks for Mariam’s many years of work developing the department and university.
Efficient streaming and enabling tomorrow's streaming services
Online video streaming has gained tremendous popularity over recent years and
currently constitutes the majority of Internet traffic. As large-scale
on-demand streaming continues to gain popularity, several important questions
and challenges remain unanswered. In his thesis, Dr. Vengatanathan
Krishnamoorthi and colleagues address open questions in the areas of efficient
content delivery for regular/linear HTTP-based Adaptive Streaming (HAS) videos
and efficient content delivery for interactive HAS videos.
In the context of regular/linear videos, Vengatanathan first investigates how HAS clients and proxy-caches can cooperate so as to improve viewers' Quality of Experience (QoE). Second, he investigates client-based prefetching techniques that download beginnings of recommended/alternative videos so as to enable instantaneous playback of these videos. Finally, the thesis presents a novel machine learning framework, called BUFFEST, which can be used by network operators to estimate clients' buffer conditions even when the transfer is encrypted and clients are streaming over HTTPS. These contributions all help improve users' everyday video streaming experiences.
In the context of interactive streaming services, Vengatanathan's thesis presents optimized solutions for two new applications: interactive branched videos and multi-video stream bundles. These solutions leverage properties of HAS to provide users with the best possible interactive user experience, allowing the viewers to influence and select the content that is being shown during playback.
Page responsible: Webmaster