[Back]

A. Frank:
"Spatial Personal Information Management: Challenges for Computer Science";
Talk: Information Systems Seminar, Oxford; 2015-02-03.

Our research focus has moved from the traditional GIS systems for large organization to personal systems we intend to use ourselves: the date in our calendars, address books, todo-list etc. contain many spatial attributes and spatial reasoning should be used for scheduling, reminders etc. This change in focus corresponds to the move of Spatial Information Management in information management from a specialty to mainstream - reflected, for example, in the ACM SIGSPATIAL successful annual conference.

The challenges we face are:

Use the application ontology for the development of the graphical user interface,
Integrate different applications and share their data, often described by ontologies,
Keep data under direct control of the person or the company,
Applications must work on-line and off-line.

To connect data available, for example, in a SPARQL endpoint to a web

applications requires too many different system, each with its own language and notation (HTML, CSS, Javascript, PHP, SQL, etc.). I assume that this complexity leads to the often frustrating experiences with web-services.

In order to facilitate integration, programming languages must allow more abstraction; second-order constructs became popular in languages like Javascript and Python (we use Haskell). Programming language should be easier to read and understand to support refactoring and restructuring of code; this helps to eventually find the best abstractions which combine without undue interactions - which, in my view, is one of the goals of computer science.

Our research focus has moved from the traditional GIS systems for large organization to personal systems we intend to use ourselves: the date in our calendars, address books, todo-list etc. contain many spatial attributes and spatial reasoning should be used for scheduling, reminders etc. This change in focus corresponds to the move of Spatial Information Management in information management from a specialty to mainstream - reflected, for example, in the ACM SIGSPATIAL successful annual conference.

The challenges we face are:

Use the application ontology for the development of the graphical user interface,
Integrate different applications and share their data, often described by ontologies,
Keep data under direct control of the person or the company,
Applications must work on-line and off-line.

To connect data available, for example, in a SPARQL endpoint to a web

applications requires too many different system, each with its own language and notation (HTML, CSS, Javascript, PHP, SQL, etc.). I assume that this complexity leads to the often frustrating experiences with web-services.

In order to facilitate integration, programming languages must allow more abstraction; second-order constructs became popular in languages like Javascript and Python (we use Haskell). Programming language should be easier to read and understand to support refactoring and restructuring of code; this helps to eventually find the best abstractions which combine without undue interactions - which, in my view, is one of the goals of computer science.

PIM, Spatial Data, Challenges