November 18, 2010: Seminar - Gianluca Quercini: "Understanding the Spatial Reader Scope of News Sources"

Seminar Announcement

Understanding the Spatial Reader Scope of News Sources

Gianluca Quercini
University of Maryland
Thursday, November 18, 2010
2:00 p.m., Room 2068 ERF


The target audience of an information source (e.g. newspaper and blog) is usually people interested in news about a specific location termed the spatial reader scope of the source. Knowledge of the spatial reader scope plays an important role in disambiguating toponyms (e.g. textual speci?cations of geographical locations) in news articles, as the process of selecting an interpretation for the toponym often reduces to one of selecting an interpretation that seems natural to those familiar with the audience location. The spatial reader scope of a news source cannot be considered as the spatial focus of the majority of the articles in the source. A ?United States? spatial reader scope of a feed is not inconsistent with the presence in the feed of many articles about ?Iraq? as readers in the United States are not just interested in what happens in the United States, but also in what can affect their lives. The key to determining the spatial reader scope of a news source is the notion of local lexicon which for a location s is a set of concepts such as, but not limited to, names of people, landmarks, and historical events, that are spatially related to s.

In the first part of this talk, which presents a joint work with Hanan Samet, Jagan Sankaranarayanan and Michael D. Lieberman, I will describe and compare techniques to automatically generate the local lexicon of a location by using the link structure of Wikipedia; in the second part I will illustrate an algorithm to determine the spatial reader scope of a news source using the notion of local lexicon; finally I will show how the knowledge of the spatial reader scope improves the geotagging of news articles.


Since June 2009 Gianluca Quercini is a faculty research assistant at the University of Maryland working with Prof. Hanan Samet. He received a MSc and PhD degree in computer science from the University of Genoa, Italy under professor Massimo Ancona. Although his PhD dissertation focused on algorithmic graph theory and graph drawing, his research interests are diverse and include virtual heritage, human-computer interfaces and, currently, information extraction and spatial databases.

Host: Professor Isabel Cruz

Copyright 2016 The Board of Trustees
of the University of
Helping Women Faculty Advance
Funded by NSF