Idiap on LinkedIn Idiap youtube channel Idiap on Twitter Idiap on Facebook
Personal tools
You are here: Home Awards and Jobs News On November 30, 2015, Maryam Habibi was awarded the EPFL doctorate for her dissertation on "Modeling Users' Information Needs in a Document Recommender for Meetings"

On November 30, 2015, Maryam Habibi was awarded the EPFL doctorate for her dissertation on "Modeling Users' Information Needs in a Document Recommender for Meetings"

Maryam Habibi arrived at Idiap as a research assistant in September 2011, for the third phase of the IM2 NCCR, after an MSc at Sharif University of Technology, Iran (2010). She prepared her thesis in the EPFL Doctoral School in Electrical Engineering (EDEE), and was a member of Idiap's NLP group. Her thesis was supervised by Prof. Hervé Bourlard and by Dr. Andrei Popescu-Belis as a co-director.

Maryam Habibi arrived at Idiap as a research assistant in September 2011, for the third phase of the IM2 NCCR, after an MSc at Sharif University of Technology, Iran (2010).  She prepared her thesis in the EPFL Doctoral School in Electrical Engineering (EDEE), and was a member of Idiap's NLP group.  Her thesis was supervised by Prof. Hervé Bourlard and by Dr. Andrei Popescu-Belis as a co-director.

In her thesis, Maryam Habibi proposed a coherent set of novel methods for extracting keywords from conversations, structuring them into topic-based queries, and ranking the obtained results.  These methods constitute a system that monitors a conversation and returns in real-time a small, relevant, and diverse set of document recommendations.  The benefits of the system were demonstrated through several comparison tasks, which were crowdsourced to a large pool of human subjects.

Maryam Habibi is now a postdoctoral researcher in the Knowledge Management in Bioinformatics group at the Humboldt University in Berlin.

We congratulate Maryam for her excellent work and for her contributions to other Idiap NLP projects, and wish her all the best for her future career.

Her thesis can be found online at http://infoscience.epfl.ch/record/213645 (EPFL Thesis n. 6760).

Dec 14, 2015
Document Actions