School of Informatics

Google PhD fellowships for Informatics students

Two students from Informatics have been awarded 2019 Google Fellowships. Nicolai Oswald received funding in the area of Systems and Networking and Reinald Kim Amplayo in Natural Language Processing.

Nicolai Oswald and Reinald Kim Amplayo

Google awards are presented to exemplary PhD students in computer science and related disciplines and acknowledge their contributions to their areas of specialty.

Addressing cache coherence protocols challenges

Nicolai is working on ProtoGen, a project that helps automate the design of cache coherence protocols. They are hardware communication protocols that keep the caches of multicore processors consistent, which are typically hard to design and verify, yet subtle bugs can seriously affect end users. Recently, a cache coherence bug caused a popular mobile phone to ship with coherence disabled, which impacted the phones’ performance and power. Initial work on this project, published last year at the International Symposium on Computer Architecture 2018, was awarded a Top Picks honourable mention by the computer architecture community.

The ProtoGen project is led by Nicolai Oswald under the supervision of Vijay Nagarajan, in collaboration with Daniel Sorin from Duke University.

The fellowship from Google will provide excellent visibility for Nicolai and ProtoGen.

Vijay Nagarajan

Taking automatic summarization to the next level

Reinald, whose work is supervised by Mirella Lapata, is interested in automatic summarization. In recent years, thanks to neural network models, automatic summarization has enjoyed renewed interest. The availability of large datasets containing hundreds of thousands of documents and their summaries has helped the development of neural architectures for summarizing single documents.

In his PhD thesis, Reinald will develop summarization models which are able to work under more challenging conditions.  He’s looking into models for generating user-tailored summaries, as well as outlines of multiple documents, and even models which learn to condense the text when document-summary pairs are not available.

I am delighted that Reinald has been selected to receive this prestigious Google PhD Fellowship. I am looking forward to working closely with Reinald and researchers at Google on representation learning for summarization.

Mirella Lapata

Fellowships include support for students as well as a Google Research Mentor. The programme was created to recognize outstanding graduate students doing exceptional work in computer science and related research areas.

 

Related links:

Google PhD Fellowship Program

Mention for ProtoGen

Nicolai Oswald’s personal page

Vijay Nagarajan’s personal page

Daniel Sorin’s personal page

Reinald Kim Amplayo’s personal page

Mirella Lapata’s personal page