Basser Seminar Series

Exemplar Queries: Knowledge Exploration Using Information Graphs

Speaker: Dr Davide Mottin
University of Trento

When: Thursday 13 August, 2015, 11am-12 noon. Please note different day and time to usual.

Where: The University of Sydney, School of IT Building, SIT Meeting Room 459, Level 4


Users often perform exploratory queries with the intent of gathering information about a specific argument. This kind of queries has an enormous potential associated with information graphs. Information graphs are graphs where nodes represent real world entities and edges relationships between them. In this work, we introduce the exemplar queries framework, a novel query paradigm that treats the user query as a representative from the set of desired results. Exemplar queries open interesting challenges and bring new opportunities for the exploitation of information graphs for exploratory search. The framework is composed of two modules. The first is exemplar query evaluation and answering, that takes in input a query and returns a set of results similar to the user query. The second performs query reformulation on the result set, and returns a set of reformulated (i.e., more specific) queries that contains, ideally non overlapping, groups of results.

The talk concludes with a thorough experimental assessment and open challenges.

Speaker's biography

Davide Mottin received his PhD in Computer Science from the University of Trento (Italy), under the supervision of Prof. Themis Palpanas and Prof. Yannis Velegrakis in 2015.

During his PhD he has been an intern in Microsoft Research Asia (Beijing, China) under the supervision of Dr. Haixun Wang, and Yahoo Labs (Barcelona, Spain) under the supervision of Dr. Francesco Bonchi.
He received the MSc and BSc in Computer Science from the University of Trento, in 2010 and 2007, respectively.

He has published on prestigious database conferences, including VLDB and KDD, and has served as an external reviewer for all major database and data mining conferences and journals.

He is a member of the dbTrento group.