Navigation auf uzh.ch

Suche

Department of Geography Geocomputation

Our research

Overview

Geographic information are ubiquitous in language, information systems and decision making. They are an important ingredient in many scientific studies, but computational methods to represent, reason with and analyse geography make many simplifications and often assume universality. Our research takes as a starting point the diversity of ways of thinking and talking about geographies, and focuses on analysis and methods to aid the use of geographic information not only by specialists, but also by society in a more general sense.

Doing so requires us to, for example, develop spatial algorithms, apply computational text analysis or do participative and ethnographic fieldwork. We work with a wide range of data sources, including traditional geographic data, social media and other forms of crowdsourced data, collections of textual data and primary data collected ex- and in-situ.

Our group is active in the field of Geographic Information Retrieval, where we have developed methods to extract and link information contained in natural language to geography, often with a focus on landscapes. Our work touches on theories of place, and we aim to make the ways in which places are understood more equitable and diverse through the use of existing and novel data sources and methods. We also develop and contribute to research on environmental processes more broadly, often with a focus on uncertainty and indicators. To do all this we collaborate widely across disciplines with other geographers, linguists, digital humanists, computer scientists and psychologists as well as industry, government and NGOs.

Thematically, our research activities encompass work on:

  • Geographic Information Retrieval
  • Digital Places
  • Landscape & Language
  • Environmental Modelling & Indicators

Geographic Information Retrieval

Does geography matter when we search for information?

We develop methods to retrieve and rank documents (e.g. texts, images) from web pages, news archives, or image collections. By linking the geographic information implicitly included in these data in Geographic Search Engines we aim to improve geographic search so that users will find the requested information more easily. Our research investigates questions like:

  • How can we effectively extract, model and utilise information about spatial relationships and vague and vernacular place names?
  • How can textual descriptions, for instance in web pages, be linked to space?

Digital Places

How can we model, represent and reason with digital representations of place?

Places are a fundamental concept in geography. However, in geographic information science they have often been abstracted to points and names or functions. In our research on digital places we explore how theories about place, develop novel representations of place-related concepts, and work on methods allowing reasoning with digital representations of place. Our research addresses questions including:

  • How can we capture and model different systems of place names, and incorporate place representations that go beyond point-like objects?
  • What aspects of diverse understandings and experiences of places can be extracted from user generated content and text?

Landscape and language

How does language influence the ways we perceive and conceptualise landscape?

Language is a gateway to understanding human interactions with the environment. What we write, read, and talk about reflects our perception. Simultaneously, language influences how we conceptualise the landscape: do people talking about Wald in German mean the same as French speakers discussing forêt. We research questions such as:

  • How can we utilise textual data to explore and understand human perception and diverse values in landscapes?
  • Are there cultural and linguistic differences in the ways in which landscapes and their elements are conceptualised - and how do these influence the ways landscapes and their management?

Environmental modelling and indicators

How can we model and monitor environmental change?

Modelling environmental change requires consideration of uncertainty and development of models capable of capturing models at appropriate scales. Reporting environmental change requires reproducible indicators, often derived from data collected for other purposes. We develop and evaluate methods combining a wide range of data sources including the biophysical (e.g. terrain data or land cover), participatory (e.g. landscape preference), user generated content (e.g. city walking tour videos) and other sensors (e.g. bicycle counters) spatial data across different spatial and temporal scales.

  • How do resolution and uncertainty in spatial data influence modelling results?
  • How can we extract information and integrate different data sources to produce useful indicators?

Bereichs-Navigation