Other Free Encyclopedias » Online Encyclopedia » Encyclopedia - Featured Articles » Contributed Topics from A-E

Context–Awareness in Mobile Commerce - INTRODUCTION, UTILIZING CONSUMER CONTEXT: OPPORTUNITY AND CHALLENGE, CONSUMER CONTEXT AND ITS CLASSIFICATION

information consumers applications aware

Jun Sun
Texas A&M University, USA

Marshall Scott Poole
Texas A&M University, USA

INTRODUCTION

Advances in wireless network and multimedia technologies enable mobile commerce (m-commerce) information service providers to know the location and surroundings of mobile consumers through GPS-enabled and camera-embedded cell phones. Context awareness has great potential for creating new service modes and improving service quality in m-commerce. To develop and implement successful context-aware applications in m-commerce, it is critical to understand the concept of the “context” of mobile consumers and how to access and utilize contextual information in an appropriate way. This article dissects the context construct along both the behavioral and physical dimensions from the perspective of mobile consumers, developing a classification scheme for various types of consumer contexts. Based on this classification scheme, it discusses three types of context-aware applications—non-interactive mode, interactive mode and community mode—and describes newly proposed applications as examples of each.

UTILIZING CONSUMER CONTEXT: OPPORTUNITY AND CHALLENGE

M-commerce gets its name from consumers’ usage of wireless handheld devices, such as cell phones or PDAs, rather than PCs as in traditional e-commerce (Mennecke & Strader, 2003). Unlike e-commerce users, m-commerce users enjoy a pervasive and ubiquitous computing environment (Lyttinen & Yoo, 2002), and therefore can be called “mobile consumers.”

A new generation of wireless handheld devices is embedded or can be connected with GPS receivers, digital cameras and other wearable sensors. Through wireless networks, mobile consumers can share information about their location, surroundings and physiological conditions with m-commerce service providers. Such information is useful in context-aware computing, which employs the collection and utilization of user context information to provide appropriate services to users (Dey, 2001; Moran & Dourish, 2001). The new multimedia framework standard, MPEG-21, describes how to adapt such digital items as user and environmental characteristics for universal multimedia access (MPEG Requirements Group, 2002). Wireless technology and multimedia standards give m-commerce great potential for creating new context-aware applications in m-commerce.

However, user context is a dynamic construct, and any given context has different meanings for different users (Greenberg, 2001). In m-commerce as well, consumer context takes on unique characteristics, due to the involvement of mobile consumers. To design and implement context-aware applications in m-commerce, it is critical to understand the nature of consumer context and the appropriate means of accessing and utilizing different types of contextual information. Also, such an understanding is essential for the identification and adaptation of context-related multimedia digital items in m-commerce.

CONSUMER CONTEXT AND ITS CLASSIFICATION

Dey, Abowd and Salber (2001) defined “context” in context-aware computing as “any information that can be used to characterize the situation of entities (i.e., whether a person, place or object) that are considered relevant to the interaction between a user and an application …” (p. 106). This definition makes it clear that context can be “any information,” but it limits context to those things relevant to the behavior of users in interacting with applications.

Most well-known context-relevant theories, such as Situated Action Theory (Suchman, 1987) and Activity Theory (Nardi, 1997), agree that “user context” is a concept inseparable from the goals or motivations implicit in user behavior. For specific users, interacting with applications is the means to their goals rather than an end in itself. User context, therefore, should be defined based on typical user behavior that is identifiable with its motivation.

According to the Merriam-Webster Collegiate Dictionary, the basic meaning of context is “a setting in which something exists or occurs.” Because the typical behavior of mobile consumers is consumer behavior, the user context in m-commerce, which we will term consumer context , is a setting in which various types of consumer behavior occur.

Need Context and Supply Context

Generally speaking, consumer behavior refers to how consumers acquire and consume goods and services (both informational and non-informational) to satisfy their needs (e.g., Soloman, 2002). Therefore, consumer behavior is, to a large extent, shaped by two basic factors: consumer needs and what is available to meet such needs. Correspondingly, consumer context can be classified conceptually into “need context” and “supply context.” A need context is composed of stimuli that can potentially arouse a consumer’s needs. A supply context is composed of resources that can potentially meet a consumer’s needs.

This behavioral classification of consumer context is based on perceptions rather than actual physical states, because the same physical context can have different meanings for different consumers. Moreover, a contextual element can be in a consumer’s need and supply contexts simultaneously. For example, the smell or sight of a restaurant may arouse a consumer’s need for a meal, while the restaurant is part of the supply context. However, it is improper to infer what a consumer needs based on his or her supply context (see below). Therefore, this conceptual differentiation of consumer contexts is important for the implementation of context-aware applications in m-commerce, which should be either need context-oriented or supply context-oriented.

The needs of a consumer at any moment are essential for determining how a context is relevant to the consumer. However, “consumer need” is both a multi-level construct and a personal issue. According to Maslow (1954), human need is a psychological construct composed of five levels: physiological, safety, social, ego and self-actualization. While it is feasible to infer some of the more basic needs of mobile consumers, including physiological and safety needs, based on relevant context information, it is almost impossible to infer other higher-level needs. Moreover, consumer need is a personal issue involving privacy concerns. Because context-aware computing should not violate the personal privacy of users by depriving them of control over their needs and priorities (Ackerman, Darrell & Weitzner, 2001), it is improper to infer a consumer’s needs solely based on his or her supply context and provide services accordingly. It is for this reason that pushing supply context information to mobile consumers based on where they are is generally unacceptable to users.

When consumers experience emergency conditions, including medical emergencies and disastrous events, they typically need help from others. Necessary services are usually acceptable to consumers when their urgent “physiological” and “safety” needs can be correctly inferred based on relevant context information. Context-aware applications can stand alert for such need contexts of consumers and provide necessary services as soon as possible when any emergencies occur. Such context-awareness in m-commerce can be denoted as need-context-awareness .

Under normal conditions, context-aware applications should let consumers determine their own needs and how certain supply contexts are relevant. The elements of supply contexts, including various sites, facilities and events, usually locate or occur in certain functionally defined areas, such as shopping plazas, tourist parks, traffic systems, sports fields and so on. Information about such contextual elements in certain areas can be gathered from suppliers and/or consumers and stored in databases. Supply-context-awareness , therefore, concerns how to select, organize and deliver such information to mobile consumers based on their locations and needs.

Internal Context, Proximate Context and Distal Context

Besides the behavioral classification, contextual elements can also be classified based on their physical locus. According to whether the contextual elements are within or outside the body of a consumer, a consumer context can be divided into internal and external contexts. An internal context is comprised of sensible body conditions that may influence a consumer’s needs. By definition, internal context is part of need context. An external context , however, can refer to both the supply context and part of the need context that is outside of a consumer.

According to whether the contextual elements can be directly perceived by a consumer, his or her external context can be divided into “proximate context” and “distal context.” A proximate context is that part of external context close enough to be directly perceivable to a consumer. A distal context is that part of external context outside the direct perception of a consumer. Mobile consumers do not need to be informed of their proximate context, but may be interested in information about their distal context. Context-aware information systems, which are able to retrieve the location-specific context information, can be a source of distal context information for mobile consumers. Besides, consumers can describe or even record information about their proximate context and share it with others through wireless network. To those who are not near the same locations, the information pertains to their distal contexts.


Figure 1 illustrates a classification scheme that combines two dimensions of consumer context, physical and behavioral. The need context covers all the internal context and part of the external context. A subset of need context that can be utilized by need context-aware applications is emergency context ; includes internal emergency context, which comprises urgent physiological conditions (e.g., abnormal heart rate, blood pressure and body temperature); and external emergency context, which emerges at the occurrence of natural and human disasters (e.g., tornado, fire and terrorist attacks). The supply context, however, is relatively more stable or predictable, and always external to a consumer. Supply context-aware applications mainly help mobile consumers obtain and share desirable supply context information. This classification scheme provides a guideline for the identification and adaptation of context-related multimedia digital items in m-commerce.

CONTEXT-AWARE APPLICATIONS IN M-COMMERCE

Context-aware applications in m-commerce are applications that obtain, utilize and/or exchange context information to provide informational and/or non-informational services to mobile consumers. They can be designed and implemented in various ways according to their orientation towards either need or supply context, and ways of collecting, handling and delivering context information.

It is generally agreed that location information of users is essential for context-aware computing (e.g., Grudin, 2001). Similarly, context-aware applications in m-commerce need the location information of mobile consumers to determine their external contexts and provide location-related services. Today’s GPS receivers can be made very small, and they can be plugged or embedded into wireless handheld devices. Therefore, it is technically feasible for context-aware applications to acquire the location information of mobile consumers. However, it is not ethically appropriate to keep track of the location of consumers all of the time because of privacy concerns. Rather, consumers should be able to determine whether and/or when to release their location information except in emergency conditions.

There can be transmission of contextual information in either direction over wireless networks between the handheld devices of mobile consumers and information systems that host context-aware applications. For applications oriented towards the internal need context, there is at least the flow of physiological and location information from the consumer to the systems. Other context-aware applications typically intend to help mobile consumers get information about their distal contexts and usually involve information flow in both directions.

In this sense, mobile consumers who use context-aware applications are communicating with either information systems or other persons (usually users) through the mediation of systems. For user-system communications, it is commonly believed that the interactivity of applications is largely about whether they empower users to exert control on the content of information they can get from the systems (e.g., Jensen, 1998). Therefore, the communications between a consumer and a context-aware system can be either non-interactive or interactive, depending on whether the consumer can actively specify and choose what context-related information they want to obtain. Accordingly, there are two modes of context-aware applications that involve communication between mobile consumers and information systems: the non-interactive mode and the interactive mode. For user-user communications, context-aware applications mediate the exchange of contextual information among mobile consumers. This represents a third mode: the community mode. This classification of context-aware applications into non-interactive, interactive and community modes is consistent with Bellotti and Edwards’ (2001) classification of context awareness into responsiveness to environment, responsiveness to people and responsiveness to the interpersonal. Below, we will discuss these modes and give an example application for each.

Non-Interactive Mode

Successful context-aware applications in m-commerce must cater to the actual needs of mobile consumers. The non-interactive mode of context-aware applications in m-commerce is oriented toward the need context of consumers: It makes assumptions about the needs that mobile consumers have in certain contexts and provides services accordingly. As mentioned above, the only contexts in which it is appropriate to assess consumer needs are certain emergency conditions. We can call non-interactive context-aware applications that provide necessary services in response to emergency contexts Wireless Emergency Services (WES). Corresponding to the internal and external emergency contexts of mobile consumers, there are two types of WES: Personal WES and Public WES.

Personal WES are applications that provide emergency services (usually medical) in response to the internal emergency contexts of mobile consumers. Such applications use bodily attached sensors (e.g., wristwatch-like sensors) to keep track of certain physiological conditions of service subscribers. Whenever a sensor detects anything abnormal, such as a seriously irregular heart rate, it will trigger the wearer’s GPS-embedded cell phone to send both location information and relevant physiological information to a relevant emergency service. The emergency service will then send an ambulance to the location and medical personnel can prepare to administer first-aid procedure based on the physiological information and medical history of the patient. The connection between the sensor and cell phone can be established through some short-distance wireless data-communication technology, such as Bluetooth.

Public WES are applications that provide necessary services (mainly informational services) to mobile consumers in response to their external emergency contexts. Such applications stand on alert for any disastrous events in the coverage areas and detect external context information through various fixed or remote sensors or reports by people in affected areas. When a disaster occurs (e.g., tornado), the Public WES systems gather the location information from the GPS-embedded cell phones of those nearby through the local transceivers. Based on user location and disaster information, the systems then give alarms to those involved (e.g., “There are tornado activities within one mile!”) and display detailed self-help information, such as evacuation routes and nearby shelters, on their cell phones.

Interactive Mode

The interactive mode of context-aware applications in m-commerce does not infer consumer needs based on contextual information, but lets consumers express their particular information requirements regarding what they need. Therefore, the interactive mode is not oriented towards the need contexts of consumers, but their supply contexts. The Information Requirement Elicitation (IRE) proposed by Sun (2003) is such an interactive context-aware application.

In the IRE approach, mobile consumers can express their needs by clicking the links on their wireless handheld devices, such as “restaurants” and “directions,” that they have pre-selected from a services inventory. Based on such requests, IRE-enabled systems obtain the relevant supply context information of the consumers, and elicit their information requirements with adaptive choice prompts (e.g., food types and transportation modes available). A choice prompt is generated based on the need expressed by a consumer, the supply context and the choice the consumer has made for the previous prompt. When the information requirements of mobile consumers are elicited to the level of specific suppliers they prefer, IRE-enabled systems give detailed supplier information, such as directions and order forms.

The IRE approach allows the consumers to specify which part of their distal supply context they want to know in detail through their interactions with information systems. It attempts to solve the problem of inconvenience in information search for mobile consumers, a key bottle neck in m-commerce. However, it requires consumers to have a clear notion of what they want.

Community Mode

The community mode of context-aware applications in m-commerce mediates contextual information exchange among a group of mobile consumers. Consumers can only share information about what is directly perceivable to them, their proximate contexts. However, the information shared about the proximate context may be interesting distal context information for others if it is relevant to their consumption needs or other interests. A group of mobile consumers in a functionally defined business area have a common supply context, and they may learn about it through sharing context information with each other. Some applications in DoCoMo in Japan have the potential to operate in the community mode.

Wireless Local Community (WLC) is an approach to facilitate the exchange of context information for a group of mobile consumers in a common supply context, such as a shopping plaza, tourist park or sports field (Sun & Poole, working paper). In such an area, mobile consumers with certain needs or interests can join a WLC to share information about their proximate supply contexts with each other (e.g., seeing a bear in a national park). Because the information shared by different consumers is about different parts of the bigger common supply context, the complementary contributions are likely to achieve an “informational synergy.” Compared with the IRE approach, the WLC approach allows mobile consumers to obtain potentially useful or interesting context information without indicating what they want.

Table 1 illustrates the primary context orientations of three modes of context-aware applications. The need context-aware applications are usually non-interactive. Personal WES applications are oriented towards the internal need context of mobile consumers, while Public WES applications are oriented towards the external (especially distal) need context of mobile consumers. The supply context-aware applications should be either of the interactive mode or community mode. As an example of interactive mode applications, IRE systems help mobile consumers know the part of their distal supply context they are interested in through choice prompts. As an example of community mode applications, WLC enables mobile consumers to share their proximate supply context with each others.

CONCLUSION

The advance in multimedia standards and network technology endows m-commerce great potential in providing mobile consumers context-aware applications. An understanding of consumer context is necessary for the development of various context-aware applications, as well as the identification and adaptation of context-related multimedia digital items. This article defines dimensions of consumer context and differentiates three modes of context-aware applications in m-commerce: the non-interactive, interactive and community modes. While applications for the interactive and community modes are in rather short supply at present, all indications are that they will burgeon as m-commerce continues to develop. Example applications are given to stimulate the thoughts on developing new applications.

Further technical and behavioral issues must be addressed before the design, implementation and operation of context-aware applications in m-commerce. Such issues may include: network bandwidth and connection, digital elements compatibility, content presentation, privacy protection, interface design, service sustainability and so on. We hope that this article can enhance further discussions in this area.

Contract on Cherry Street [next] [back] Context-Aware Video Production - Context Awareness, Context Enhanced Retrieval of Video, Vision

User Comments

Your email address will be altered so spam harvesting bots can't read it easily.
Hide my email completely instead?

Cancel or