Location-Based Web Services for Car Infotainment Susanne Boll - - PowerPoint PPT Presentation

Location-Based Web Services for Car Infotainment

Susanne Boll

University of Oldenburg & OFFIS Institute for Information Technology

01.07.2009 1st tubs.CITY Symposium, Braunschweig

52°16‘30"N 10°31‘43"E

C3World projekt partners

Niedersächsisches Ministerium für Wissenschaft und Kultur OFFIS Oldenburg - Institut für Informatik Technische Universität Carolo Wilhelmina zu Braunschweig Leibniz Universität Hannover Volkswagen Konzernforschung

Overview

Grounding and Mapping of Location and Content In-Car Usage Analysis of Web Content

Content and Location External Sources, Maps, Indexes Navigation and Infotainment

Motivation

• Location-related information need
• Users require location-related

information

• Up to 20% location-related queries

at general search engines

• Spatial information on the Web
• Estimated location references in

up to 20% of Web pages

• Richer information than most

commercial databases

• Valuable source of location-based

information

Geographic Web Information Retrieval

• Matching of Web information

to the real world

• Semantic retrieval of location

information

• Identification, extraction and

processing of geographic references

The challenge

• Harvesting, identifying, grounding,

augmenting, exploiting, visualizing location information

• Geospatial information on the Web is:
• Hidden in Web pages
• Metadata: Geographic coordinates
• Content: Addresses, Place names
• Scattered
• Present in a variety of different non-

connected pages

• Not explicitly structured
• Mostly written in the textual content

with typos, omission etc. without known relation to content

<META name="geo.position" content="53.1467;8.2154">

Geospatial Web Search

• Discovery, extraction, processing
• f location references
• Classification of location-relevant

content

• Semantic enrichment of location

and content

• Hybrid search on geospatial,

textual and further semantic aspects

Coordinate: 53.146712°N 8.215467°E

Geospatial Retrieval Processes

discover understand augment explore

crawling geoparsing link analysis identification mining scope analysis semantic enrichment location extension visualization interaction feedback integration annotation

Spatial search as new application for in-car infotainment

• Understanding of spatial

characteristics

• Efficient geospatial search engine
• In-car Search
• Special (mobile) requirements and

information needs

• Information for the here and now
• Development towards integrated

high-value services

Requirements of car-based search

• Crawling, indexing and querying of Web

content

• Location- and route-specific content
• Information for drivers and passengers
• Search interaction
• Spontaneous and ‚standing‘ search
• Visualization within the car systems
• Preprocessing and presentation
• Integration
• Communication of car and search engine
• Context processing
• Connection to car and navigation systems

discover understand augment explore

Architecture

Spatial Web Index Query Engine Webcrawler, Indexer, Geoparser Context

In-Car System

Car-To-X

Search Engine

Presentation, Visualization, Interaction Preprocessing, Context Processing

Context in the car

• Context for improvement of search results

and situation-aware presentation

• Adaptation to situation and information need
• Spatial context
• Position, route, destination,…
• Additional context available in the car
• CanBUS
• Duration of trip
• User profile
• Familiarity with route
• Type of trip
• …

In-Car Context Automotive Context Spatial Context User Context

Spatial context is more than position

• Route estimation, departure,

destination, speed, time, distances, …

• Support of mobile users‘

information needs

• Result adaptation in mobile IR
• Spatial Query + Spatial Filters +

Spatial Ranking Spatial Search

• Hybrid search: combination of

spatial and textual features

Interface and Presentation

• Challenges
• Restricted interaction
• Restricted attention
• Fast mobile movement
• User interaction
• Intuitive query formulation and

handling

• Use of rich context information
• Dynamic interaction and exploration
• Presentation of content
• Preprocessing and adaptive

visualization

• Use of semantic aspects to match query

and content

Conclusion

• Spatial search as new application for

in-car infotainment

• Spatial search engines for trip-based

and roadside information

• Navigation as ‚smart passenger‘
• Research challenges in
• Web search
• Geographic information retrieval
• HMI
• Automotive Infotainment

Q&A

Susanne Boll Media Informatics and Multimedia Systems University of Oldenburg, Oldenburg, Germany susanne.boll@informatik.uni-oldenburg.de 53° 8′ 55.9″ N 8° 12′ 0.43″ E