2. Mobile Travel Guide Information Systems
 
Starting in the 1990s (Norman 1990) scientists realized that the way people interact with computers is unnatural and inefficient. The problems people are having with computers are similar to the problems they have with maps. To use paper maps people have to put themselves mentally into the map’s frame of reference. Research in the ubiquitous computing domain (Weiser 1993) has shown that one way of making computers easier to use is by providing the computer system with some awareness of the user’s context (Schilt et al. 1994). Context awareness allows the system to have insights into the user’s frame of reference and can provide information that is aligned with this frame (Kuipers 1982); therefore, the user’s interaction with the computer system seems more intuitive. This idea of context awareness is useful in the spatial information domain. If a map is aware of its user’s spatial context then the map can be centered on and oriented around the user, thereby reducing the spatial cognitive load for people (Schilit and Theimer 1994).

In 1991, Mark Weiser introduced the idea that the arcane aura that surrounds personal computers is not just a “user interface” problem, it is that the idea of a “personal” computer itself is misplaced (Weiser 1991). To fix this problem he suggested that computing become more distributed and ubiquitous. A goal of ubiquitous computing is to make users less aware of computing devices so that the information the computer supplies is easier for users to understand. In order for this to happen the computer devices need to disappear from the user’s consciousness. This conscious disappearance is a necessary fundamental consequence of human psychology, not of technology. Whenever people learn something sufficiently well, they cease to be aware of it (Norman 1999).

In order for a computer system to be ubiquitous it needs to dynamically adapt to the environment (Harter et al. 1999). A computer can intelligently adapt if it is aware of some contextual information about its usage, information about the user, the user’s environment, and time (Rodden et al. 1998). The necessity of this contextual information shows how context-aware computing was conceived out of ubiquitous and mobile computing. Spatial awareness is a fundamental component of context-awareness (Dix et al. 2000), for example, in order for a personal digital assistant to become aware that its user is making coffee, it would need to know that its user is in front of the coffee machine and maybe that it is morning. This example shows that in order to understand why and how some event is occurring one first needs to know when and where this event is happening.

To understand how a map can become spatially-aware, an investigation into other spatially-aware mapping systems is conducted, showing the tasks they perform, the manner in which they collect and use spatial data, and how users interact with these systems. After an examination of related spatially-aware geographic information projects is performed an investigation of augmented reality system is presented. This augmented reality investigation shows how these systems exploit many similar technologies that spatially-aware mapping systems have, thereby showing that the egocentric spatial data model developed in this research is useful for augmented systems as well. This chapter concludes with some background information about the standard query language and how database management systems can be extended with abstract data types. A description of current spatial and temporal database system is given, which is necessary because the egocentric datatype inherits properties from these current database extensions. This backdrop information is also useful to gain an understanding of the egocentric spatial data model that is developed in Chapters 4 and 5.

2.1. Spatially-Aware GISs

Spatially-aware GISs encompass more than the standard idea of GIS. Travel aids should also be considered GIS because they possess mapping and wayfinding assistance as well as connecting spatial data with attribute data. Digital travel guides have provided some of the most innovative concepts for spatial awareness. To provide spatial information for wayfinding processes and increase users’ interactions with the environment, Spatially-aware GISs use sensor-enriched mobile computing devices. These devices do not just provide information about a spatial entity, but also connect that information to other background information, such as the entity’s history or its physical properties. These GISs need to collect and process relevant information in extremely variable environments.

The ground-breaking Lancaster GUIDE enables visitors to flexibly explore and learn about a city in their own ways (Cheverst et al. 2000). The system is capable of acting as an intelligent tour guide or as a richly featured guidebook, depending on the visitor’s needs. GUIDE also allows visitors to control their pace of interacting with the system and the environment. At every popular travel attraction an 802.11 base station is located for wireless data transmission. The spatial information in this situation is very coarse, for example, at the Lancaster Castle it might be possible to tell on which side the user is. Since the sensed positional data, which is based on wireless triangulation, is imprecise, the system’s feedback must also be at a coarse granularity. The GUIDE has to provide coarse information because the system cannot assume a precise contextual relationship between the user and geographic space. GUIDE can provide information about a large area, like the north side of the Lancaster Castle.

The GUIDE prototype uses Tablet computers, which allows for varied forms of user interaction. The system acts like a spatially driven web browser where users are presented with text, pictures, sound, and maps. Users are also informed of other people using the system nearby so they could send messages to each other about their experiences.

The Outdoor Cyberguide (Abowd et al. 1997) functions as a digital guidebook on a mobile PDA, not an augmented reality system. It provides text and maps for users to interact with, but does not provide pictures or sound. It also allows users to send instant messages to other current tourists. The Cyberguide system is unable to change the underlying maps automatically.

The Deep Map Framework (Zipf and Aras 2002) includes a web-interface for pre-trip planning, language-driven mobile prototypes on a portable computer as well as location-based services for smartphones. Its services are intelligent position determination, the dynamic generation of individualized proposals for sightseeing tours and adaptive maps. Deep Map is not an augmentmented reality environment as the main human interface is a Pocket PC device. Instead, users read 2D maps, background text, as well as look at a 3D model of the environment with augment information. This system is voice-driven and the presentation information is based on spatial proximity.

The Tourism Information Provider (Hinze and Voisard 2003) delivers various types of information to mobile devices based on location, time, profile of end users, and their “history”. The TIP notification system is composed of mobile devices and a server. The server is comprised of three thematic databases; profile, scheduled event, and spatial. The TIP system is a nice example for showcasing most of the information necessary for an intelligent mobile GIS. A problem that the TIP has that is addressed by our egocentric spatial data model is its data management is dispersed where interoperability problems arise. The egocentric spatial data model incorporates these separate databases into a single instance of a database management system. Placing the profile, scheduled, history, and spatial data into a single instance allow the system to be autonomous, consistent, and durable.

Another type of spatial information systems are augmented reality systems. These systems overlay spatial features within a user perspective view of their surroundings with attribute information. Though augmented reality systems supply a different form of user interaction than mobile GISs due many of the functionality is similar. Users of augmented system still need spatially-aware information and an ability of egocentrically query about their surroundings.

2.2. Augmented Reality Systems

Augmented reality systems use spatially-aware sensors to help users by augmenting the user’s view of reality with geospatial information. The “map-in-the-hat” (Thomas et al. 1998) enables users to see waypoints and Compass headings for walking directions through a head-mounted display. The focus of the “map-in-the-hat” is to guide users in an orienteering task from one point to another. In the “map-in-the-hat” system, users see attributes, but they are not presented with a map to read.

Battuta (Nusser et al. 2003) seeks to enable access and use of digital geospatial information for field data gatherers who do not have extensive training in spatial analysis. The Battuta prototype is a wearable system that includes a digital compass and a GPS receiver. The augmented data is displays through a viewer clipped to a regular pair of glasses. The augmented vision has a see-through map plus a locator for the users to read while performing their data gathering tasks.

The Touring Machine (Feiner et al. 1997) is a 3D mobile augmented reality system for exploring the urban environment. The prototype assists users by overlaying information about items of interest in their vicinity. As a user moves about, he or she is tracked through a combination of GPS positional and magnetometer/inclinometer orientation sensors. Information is presented and manipulated on a combination of a see-through, headworn 3D display, and an untracked, opaque, handheld 2D display with stylus and trackpad. This system allows the user to continually see a building in reality, read a label of the building, read a 2D map, and read background text.

The next section examines the standard relational and object-relational database systems, and how database management systems have been extended for spatial and temporal data. This analysis is provided as background information for the development of the egocentric spatial data model, which is a database extension that inherits properties from the spatial and temporal data extensions.

2.3. Using and Extending Database Management Systems

Database query languages are tools to facilitate the access to a database and have been investigated in computer science for many decades (McFadden and Hoffer 1985). We use the term query for a statement requesting the retrieval of data from a database. For the description of queries the Structured Query Language (SQL) is used (Chamberlin and Boyce 1974). SQL is the standard relational query language and enjoys popularity in traditional database applications.

2.3.1. Structured Query Language

The fundamental structure of SQL is the SELECT-FROM-WHERE block. The SELECT clause determines the attributes to display, the FROM clause describes the data sets needed to solve the query and the optional WHERE clause specifies constraints upon the items to be retrieved (Ramakrishnan and Gehrke 2000). Here is an example of a common SQL query:

> SELECT    population
> FROM    town
> WHERE    name = “Orono”;

2.3.2. Spatial Data Queries

Spatial SQL commands (Egenhofer 1994) for the selection of spatial data improve upon the fundamental SQL block by adding support for spatial predicates, such as topological features and object geometries for points, lines, and regions (Guting 1994). An example of a Spatial SQL query is:

> SELECT    lake.name
> FROM    lake, city
> WHERE    lake.geometry north_of city.geometry
> AND    city.name = “Orono”;

In this example lake is the data type, the term north_of is the spatial predicate, and Orono is the reference item. This query uses the abstract data type geometries, which incorporates spatial relations, such as inside, covered by, overlaps, equals, contains, and covers.

2.3.3. Abstract Data Types

In order to serve as support for non-built-in types such as; spatial, temporal, and egocentric data the relational data model has been extended. Such data types and their operations are referred to as Abstract Data Types (ADTs) (Guting 1994). ADTs were introduced as a way to circumvent the lack of modeling power inherent in the basic relational database model (Stonebraker 1986). Within extended relational systems, users manipulate values through the use of reference queries whose types are basic, such as integer or real types, but also abstract data types accessible through the operations defined on them as methods.

The term ‘abstract’ is applied to these data type because the database system does not need to know how an ADT’s data is represented nor how the ADT’s methods work. It merely needs to know what methods are available and the input and output types for the methods. Hiding of ADT internals is called encapsulation. In an object-relational system, the simplification due to encapsulation is critical because it hides any substantive distinctions between data types and allows object-relation database management systems to be implemented without anticipating the types and methods that users might want to add.

2.3.4. The Spatial Data Type

The term spatial data is used in a broad sense, covering multidimensional points, lines, rectangles, polygons, cubes, and other geometric objects (Guting 1994). A spatial data object occupies a certain region of space called its spatial extent, which is characterized by its location and boundary. From the point of view of the database management system spatial data is classified as either point or region data. Queries that arise over spatial data are of three main types: spatial range queries, nearest neighbor queries, and spatial join queries.

2.3.5. The Temporal Data Type

Time is an important aspect of all real-world phenomena. Events occur at specific points in time; objects and the relationships among objects exist over time. The ability to model this temporal dimension of the real world and to respond within time constraints to changes in the real world as well as to application-dependent operations is essential to many computer applications (Snodgrass 1995).

Discrete interpretation of time has commonly been adapted by the research community in temporal databases because of the simplicity and relative ease of implementation (Tansel et al. 1993). Hence, time will be interpreted as a set of equally spaced and ordered time points and denote it by T. T = {0,1,2,..now…}. Any point beyond new is future time. Any interval or temporal element that includes the special constant now expands as the value of now advances.

The egocentric spatial data model inherits properties from both the spatial and temporal database management systems in order to process the terms here and there. In order for a database to process here and there queries it needs to be aware that these queries are geospatial and occurring now. For a database to allow queries where a user is cognitively immersed in the spatial data environment the egocentric spatial data model is necessary as shown in Chapters 4 and 5.

2.4. Summary

This Chapter examined related projects, comprised of ubiquitous and spatially-aware computing. The descriptions showed that there are systems that use sensors such as GPS as well as orientation sensors. The goal of this examination was to see if a map can be improved by giving it an insight into the user’s needs and which form he or she wants the information to be presented. A key aspect missing from the projects described in this chapter is a robust query environment of mobile, distributed, and egocentric information systems. It is because of this lack of a query environment that the second half of this Chapter provides some background about query languages and other common data models. This is necessary before the development of the egocentric spatial data model in Chapters 4 and 5, which show how the spatial data query language should be extended for egocentric spatial queries.