Journal of Geographic Information System, 2012, 4, 425-443 Published Online October 2012 (
Mobile and Context-Aware GeoBI Applications: A
Multilevel Model for Structuring and Sharing of
Contextual Information
Belko Abdoul Aziz Diallo, Thierry Badard, Frédéric Hubert, Sylvie Daniel
Centre de Recherche en Géomatique, Université Laval, Québec, Canada
Received February 27, 2012; revised March 26, 2012; accepted April 23, 2012
With the requirements for high performance results in the today’s mobile, global, highly competitive, and technol-
ogy-based business world, business professionals have to get supported by convenient mobile decision support systems
(DSS). To give an improved support to mobile business professionals, it is necessary to go further than just allowing a
simple remote access to a Business Intelligence platform. In this paper, the need for actual context-aware mobile Geospa-
tial Business Intelligence (GeoBI) systems that can help capture, filter, organize and structure the user’s mobile context is
exposed and justified. Furthermore, since capturing, structuring, and modeling mobile contextual information is still a
research issue, a wide inventory of existing research work on context and mobile context is provided. Then, step by step,
we methodologically identify relevant contextual information to capture for mobility purposes as well as for BI needs,
organize them into context-dimensions, and build a hierarchical mobile GeoBI context model which 1) is geo-spa-
tial-extended; 2) fits with human perception of mobility; 3) takes into account the local context interactions and informa-
tion-sharing with remote contexts; and 4) matches with the usual hierarchical aggregated structure of BI data.
Keywords: Context-Awareness; Decision Support System (DSS); Mobile Geospatial Business Intelligence (GeoBI);
Decision-Making; Relevant Contextual Information; Context Dimensions; Context Modeling; Context
Sharing; Context Structuring; BI Data
1. Introduction
In the today’s global, highly competitive and technology-
based business world, business professionals are not only
becoming increasingly mobile, moving to places where
business requires them (opportunities to catch, problems
to deal with, meetings to attend, etc.), but they also have
to keep looking after the fluctuations and trends of their
businesses at anytime from anywhere, via mobile appli-
cations and devices (smart phones, PC pockets, etc.), in
order to be able to take the right decision at the right time
and hence to reach their business goals. To assist these
offsite business people, suitable mobile decision support
systems (DSS) are required.
Thanks to the emergence of pervasive and mobile
computing technologies, several mobile DSSs have been
proposed amongst: Hand-OLAP [1], Mobile-OLAP [2],
Spatial OLAP Mobile [3], MoBI [4], and Go! Mobile [5].
These solutions are actually mobile Business Intelligence
Systems (BIS) built upon data warehouse and OLAP
technologies. Business Intelligence Systems (BIS) has
emerged from the 1990’s [6] as convenient means to
collect, store and compute the transactional data soup1
into summarized and meaningful information for deci-
sion makers.
Based on cross-readings of different authors and prac-
titioners ([7-15]), we define Business intelligence (BI) as
the activity or the process of intelligently gathering, in-
tegrating, aggregating, storing, processing and analyz-
ing business data in order to extract or find out synthe-
sized, pertinent and meaningful information and knowl-
edge in a way that improves business decision mak-
ing. Criteria specifying what is relevant and meaningful
for an organization are often predefined and measured
through metrics and key performance indicators (KPI)
which can be displayed in dashboards, scorecards, re-
ports, etc., as crosstabs, di a grams, and m ap s.
Business Intelligence Systems (BIS) refer then, to the
adequate computer-based tools and technologies (e.g.
Extract, Transform and Load (ETL) tools, data ware-
1Data soup is an expression introduced by [56] to emphasis the huge
amount of detailed and raw data scattered in multiple various data
sources resulting from mass computerization and the advent of infor-
mation and communication Technologies (ICT).
opyright © 2012 SciRes. JGIS
housing, Online Analytical ProcessingOLAPdata min-
ing, reporting, dashboarding, etc.) which, when harmo-
niously arranged, can support the implementation of the
analytical process in a comprehensive, reliable and ef-
fective way, in order to assist decision makers in moni-
toring and analyzing their businesses.
Since business intelligence and mobility are con-
cerned, geospatial features are obviously involved in,
such as: international or national borders (e.g. continents,
countries, provinces, etc.); transportation infrastructures
(roads, railways, planes, trains, etc.); place names, ad-
dresses or zip codes (of sales, deliveries, meetings, cli-
ents, partners, companies, services, etc.); GPS coordi-
nates of geo-localized POIs, routes and directions to fol-
low, etc. According to [16], about 80% of data store or
corporate data warehouse has a geo-spatial dimension.
That geo-spatial part of business data may be more suita-
bly exploited if analyzed and represented by BI Systems
coupled with maps and GIS capabilities. This geospa-
tial-extended business intelligence is known as Geospa-
tial Business Intelligence (GeoBI). Furthermore, GeoBI
systems, by providing an intelligent coupling of geospa-
tial and Business Intelligence technologies have specifi-
cally extended the capacity of data analysis by bringing
spatio-temporal data support, cartographic visualization
and spatial analysis capabilities. Except Spatial OLAP
Mobile [3], the mobile DSSs aforementioned are not
Considering mobility as the fact of getting far (re-
moteness) from the organization’s resources (humans,
data, work tools, etc.) and losing a certain direct contact
with them, all the mobile BI Systems listed previously
have then been primarily designed to provide mobile
decision makers with access from anywhere at any time
to their business data and analysis tools, remotely as if
they were in their office: no matters where the user is and
what is around him. Therefore, if they conveniently pal-
liate to the need of data access and analysis once on the
ground ([14]), they do not really exploit the user’s loca-
tion, and ignore his dynamic changing work context
where some events may influence or improve his initial
insights (e.g. traffic congestion, available services, wea-
ther conditions, noise, people met, local security, busi-
ness or administrative alerts, etc.).
To give an improved support to mobile business pro-
fessionals, it is necessary to go further than just allow-
ing a simple remote access to a Business Intelligence
platform. An actual context-aware mobile Geospatial
Business Intelligence (GeoBI) system that fully takes
into account all aspects of mobility is required. Indeed,
we believe that in addition to being BI-based, a suit-
able mobile DSS should be, for well-informed and en-
vironment-adapted decisions: 1) geospatial-extended, i.e.
GeoBI-based, to integrate geospatial aspects of business
and mobility; and 2) context-based, to capture the user’s
reality in mobility.
As it is known, professional mobility is usually justi-
fied by the mobile worker’s need to experiment a physic-
cal proximity [17] to resources, problems or opportuni-
ties such as meeting persons (customers, suppliers, part-
ners, etc.), scrutinizing more closely some phenomena
(pollution, traffic, sales on the field, etc.), visiting POIs,
etc., in order, among other things, to acquire more accu-
rate and precise information from what he sees, hears,
smells, feels, or senses within his local work context, and
in consideration of influences of remote contexts. This
contextual information may be of any kind and related to
business, social, environmental, geospatial or techno-
logical issues so that some part of the information may
escape the mobile worker’s attention while the another
part may ask for more cognition effort to get managed
and understood. A context-aware mobile GeoBI system
should help capture, filter, organize and structure it into a
human perception-compatible context model so that it
would be exploited in combination with BI data struc-
tures and models (data warehouses, data cubes, etc.) to
provide the users with appropriate analytics on which
they can base their decision process and take fully in-
formed decisions.
Modeling and structuring mobile contextual informa-
tion into suitable context models is still a research issue,
mainly for context-based mobile GeoBI solutions. The
wide inventory of existing research work on context and
mobile context we provide in the first part of this paper
shows that there is still neither a model nor an inventory
about relevant contextual information for mobile GeoBI
contexts. In the rest of the paper, we identify relevant
contextual information for mobility issues as well as for
BI needs and organize them into context-dimensions.
Then, we propose and build step by step, a hierarchical
mobile GeoBI Context Model (named GeoMoBICoMod)
which 1) is geo-spatially-extended; 2) fits with human
perception of mobility; 3) takes into account the local
context interactions and information-sharing with some
remote contexts; and 4) matches with the usual hierar-
chical aggregated structure of BI data.
2. Related Works on Context and Mobile
The notion of context is widespread in manifolds do-
mains, but is often differently apprehended and defined
even in the same area of research. Therefore, there is still
no consensus among authors on what should be a con-
2.1. Context Definitions
Current definitions of context are ranging from human
Copyright © 2012 SciRes. JGIS
experience to robotics. For instance, considering the
user’s experience, Bolchini et al. [18] defined context as
an active process dealing with the way humans weave
their experience within their whole environment, to give
it meaning.”; whereas in robotic vision, Lombardi et al.
[19] considered that “context is what imposes changes to
the variable part of a system, (…) a particular configu-
ration of internal parameters”.
From a mobile computing standpoint in which we are
interested in, Chen and Kotz [20] adopted a software-
based definition: “Context is the set of environmental
states and settings that either determines an applica-
tions behavior or in which an application event occurs
and is interesting to the user”. This technology-oriented
definition implicitly refers to the concept of con-
text-awareness of applications.
2.2. Context-Awareness
The notion of context-awareness originates from Schilit
et al. [21] who introduced context-aware applications as
software that “adapts according to the location of use,
the collection of nearby people, hosts, and accessible
devices, as well as to changes to such things over time”.
From the viewpoint of several works, context awareness
refers to the ability of an application or a device to adapt
itself to its environment, 1) automatically (active aware-
ness); 2) at the user request (passive awareness); or 3)
based on the user’s preferences (personalization and
adaptability, i.e. the possibility for users to configure
how the application should behave according to the con-
text) [22,23]. According to Dey [24] “a system is con-
text-aware if it uses context to provide relevant informa-
tion and/or services to the us er, where relevancy depends
on the users task.
History records that the Olivetti Research group and
Xerox Parc group, with their pioneering work respect-
tively on active badge location systems [25] and ubiqui-
tous computing experiment [26], were the firsts to issue
context-aware systems. In the case of active badges [25],
the authors proposed the use of infrared active badge for
a direct localization of office staff. The ubiquitous com-
puting experiment [26] was about locating and displaying
people—their faces—on an indoor dynamic map (rooms)
to follow people’s activity. Over the time, research work
on context and context-awareness has increased tenfold
and embraces various domains such as context-based
business activities (e.g. [27,28]), mobile work (e.g. [29,
30]), mobile web ([31]), mobile mapping (e.g. [32,33]),
and mobile context-aware computing in general (e.g.
[34,35]), etc. But there is still not yet works on con-
text-based GeoBI to our knowledge.
To be context-aware, applications have to sense and/or
be aware of contextual information identified as relevant
for the targeted purpose. In the literature, some contex-
tual information has been inventoried by different au-
thors as being part of context.
2.3. Context Content: Major Contextual
Information and Context Dimensions
in the Literature
Several works have proposed a set of contextual infor-
mation a context might contain. Context content can be
viewed as the whole set of contextual information in-
volved in that context. According to Dey et al. [36], con-
textual information is any information which can charac-
terize the situation of a relevant entity (e.g. the user) for
the application. Moreover, the work of Winograd [37] on
context architectures highlighted that it might be a dif-
ference between contextual information, and the envi-
ronment settings: “something is context because of the
way it is used in interpretation, not due to its inherent
properties”. Thanks to the survey of Chen and Kotz [20]
on “context-aware mobile computing”, it is also known
that a context content (or a part) might be active “that
influences the behaviors of an application”, or passive
context (e.g. environment settings).
In the literature, beyond these different natures of
contextual information, context content is generally de-
termined according to the context-aware application
purpose and its design standpoint. Several authors have
proposed and listed contextual information they counted
as relevant, generally by organizing it into models or
context dimensions, i.e. main categories of contextual
information (e.g. social context, time context). These
main categories are generally described with key ele-
ments or sub-categories (e.g. weather, seasons for time
context) depending on the standpoint and the targeted
activity (purpose) authors considered. For example,
while [36,38] considered context in general as regarding
location, people and objects, [30] organized it, in the
case of mobile work, into five main dimensions: Task
context (interactions with the system, tasks related to the
work); Social context (people, work community, culture);
Infrastructural context (technologies, device, system,
etc.); Spatial context (location, temperature, noise, etc.);
and Temporal context (schedules, deadlines, etc.).
Regarding context-based business activities, some few
works are emerging. Those we found are mainly oriented
to online activities. The first one [27] proposed a “con-
text model for B2B collaborations” for online exchanges.
The paper identified three main dimensions with several
sub-dimensions or key elements: User/Company (User
expression, Profile, Industry Sector, Product Service,
etc.); Temporal context (Time Expression, Periodicity,
Lead time, etc.); and Location (Geographical Expression,
Copyright © 2012 SciRes. JGIS
Transportation Mode, Political Stability Index, Tariffs,
etc.). Interested in context-based e-commerce, the second
one [28] defined such context as a combination of two
main dimensions (user context and business context)
characterized by other context dimensions as follow:
User context (Personal context, Task context, En-
vironmental context, Social context).
Business context (Product Context [product category,
price, features], Business rules).
Tackling context according to human spatial cognition,
[39] and [40] elaborated context into three hierarchical
From the viewpoint of [39], human beings think about
the real space by dividing it into hierarchical mental
The body space which is the mental representation of
the body shape and postures through body’s junctions
(head, arm, hand, chest, back, leg, foot) and senses:
eyes, ears, mouth, etc.;
The space around the body which is the space of
things that can be seen or reached. It is referred to
through three main axes in a 3D frame: head/feet,
front/back, and left/right axes;
The navigation space which is the less known and
sensed space with simplified or aggregated spatial
information referred to by the means of landmarks,
paths, links and nodes.
Considering the perception of life space, [40] have
also identified three hierarchical spaces, but meaning-
fully different from those identified by [39]:
The vista space which is the space around the body
where activities are regularly carried out such as
home or work place;
The local displacement reinforcement space that is
the frequently visited space (usually by feet) around
the vista space;
The enlarged displacement reinforcement space which
consists of the region that embraces both the activity
islands beyond the local displacement reinforcement
region and the local reinforcement region itself.
In the case of mobile context-aware computing, [34]
and [35] similarly organize context content into five di-
mensions, but with some different dimensions. The simi-
lar dimensions are:
User context (username for [34]; goals, tasks, inten-
tions, etc. for [35]);
Location context (position, GPS-coords) for [34] or
Physical context (objects) for [35];
Time context (current time, system clock) for [34] or
Temporal context (time) for [35].
The rest of dimensions are different. While [34] iden-
tified Device context (device identifier, device type) and
Network context (network connection types, bandwidth)
for the purpose of making remote devices communicate,
[35] proposed Social context (people) and Computing
context (connectivity, network capacity, processors, etc.)
as relevant in the purpose of multidisciplinary context-
aware computing.
Besides, it can be noticed that in addition to structur-
ing contextual information into dimensions, [34] has ad-
ditionally separated contextual information related to
mobile devices into local context (“the context of the
location device”) and remote context (“the context of
remote devices”). This location-based contextualization
of contextual information is relevant for us and will be
exploited and extended to mobile geospatial business
intelligence aspects later in Section 4.2 to build step by
step a hierarchical and multilevel model for mobile
GeoBI contexts.
Other main works proposing some relevant context
contents are reviewed in Table 1 which provides in fact,
a summary of a somehow wide inventory. The criteria of
context-awareness, mobility, and enough detailed de-
scription of contextual information organized or not into
models have been mainly considered in addition to
originality (specific viewpoint different from others).
Table 1 presents contextual information listed by authors.
Contextual information is presented in the form of the-
matic-based groups known as dimensions (bold, italic,
and underlined terms in the table), each dimension con-
taining (between brackets in the table) some key and de-
tailed contextual elements identified by authors, if any.
From this large inventory of work on context and mo-
bile context, it can be noticed that there is still neither a
model nor an inventory about relevant contextual infor-
mation and dimensions for a mobile Business Intelli-
gence Context. Works of [27,28], even if related to busi-
ness activity, are not dedicated to mobile activities, and
seem to be too restrictive and not easily adaptable and
expandable to mobile business intelligence, according to
our viewpoint. For instance, they do not integrate the fact
that with the globalization of economy, a given local
business context (e.g. Hong Kong stock exchanges) may
be influenced by a remote one (e.g. Wall Street stock
exchanges). In addition, the lack of contextual informa-
tion in their business contexts about markets, resources,
and strategies (e.g. objectives, actions plan, metrics, etc.)
regarding organizations (government, companies, etc.)
means that a deep reorganization would be required to
extend them for business intelligence purpose.
It can also be observed that works on context are
mainly focused on applications’ context-awareness. But,
since users are also sensitive to events occurring in a
mobile context (this is known as situation awareness), it
should be relevant to consider how applications’ context-
awareness and users’ situation-awareness could be con-
nected in order to substantially enhance decision-making
in mobile environments.
Copyright © 2012 SciRes. JGIS
Copyright © 2012 SciRes. JGIS
Table 1. Context dimensions and key elements in the lite r ature .
Standpoint Purpose Authors
Context content: Context dimensions and their key and detailed elements listed by
Context in general Any activity [36,38] -Location; -People (identity, state); -Objects (computational and physical);
[41] -Human factors (user, social environment, tasks); -Physical environment
(Conditions [light, pressure, acceleration, temperature], infrastructures, location);
Local and Remote contexts composed of:
-Time context (current time, system clock); -Location context (position,
-Device context (device identifier, device type); -User context (user name);
-Network context (network connection types, bandwidth);
-User context (goals, tasks, intentions, history, preferences); -Physical context (objects);
-Social context (people); -Temporal context (time);
-Computing context (connectivity, network capacity, computing costs, display &
input, processors);
User context divided into:
-Environmental context (users surroundings: things, services, light, people, information
accessed by the user); -Personal context ( the mental and physical information about the
user: mood, expertise, disabilities, weight); -Social context (friends, relatives, colleagues);
-Task context (user’s goals, tasks, activities, etc.); -Spatio-temporal context (time, location,
-Internal context (user state, experience, user goals, tasks, current projects, status, to-do
items, personal events, user’s cognitive state, user’s emotional state and physical state (e.g.,
-External context (temperature, time, location, people, devices, etc.);
Any mobile
Active context vs. Passive context containing:
-Computing context (network connectivity, communication costs, communication
bandwidth, printers, displays, workstations, etc.); -User context (user’s profile, location,
people nearby, social situation); -Physical context (lighting, noise levels, traffic conditions,
-Time context (time of a day, week, month, and season of the year);
Mobile active
map service [32] -Location; -Time; -People; -Objects (printers, terminals, workstations, etc.);
-Services (location-based services);
-Computing system context (display size, network connectivity, communication costs and
bandwidth, nearby resources (printers, displays), etc.); -User context (user’s profile and
-Social context (people nearby); -Cultural context (characters, date and time formats);
-Physical context (physical surroundings [Lighting, temperature, weather conditions, noise
levels], location, orientation); -Time context (time of day, week, season of year); -History
context (spatial navigation history);
mobile mapping
LBS [33]
-User preferences (dietary restriction, range of price, and acceptable restaurant rating);
-User context (location, available time, and privacy requirements); Environmental context
(e.g., time, weather, other user reviews, and current traffic); -Database-specific context
(e.g., for a restaurant, considering current waiting line, opening status, rating, and change of
[29] Cited
by [30] -Mobile context; -Mobile workers; -Mobile technologies; -Mobile tasks;
Mobile work Any mobile
work [30]
-Task context (interactions with the system, tasks related to the work); -Social context
(people, work community, culture); -Infrastructural context (technologies, network
connections, device, system, service ecosystems); -Spatial context (place, location,
temperature, noise, lighting, furniture); -Temporal context (schedules, deadlines, place of
work, (ir)regularity, planned/unplanned, day time or week);
User experience Any mobile
work [45]
-User (values, emotions, expectations, prior experiences, physical characteristics, motor
functions, personality, motivation, skills, age, etc.); -Social factors (time pressure, pressure
of success and fail, explicit and implicit requirements, etc.); -Culutal factors (sex, fashion,
habits, norms, language, symbols, religion, etc.); -Context of use (time, place,
accompanying persons, temperature, etc.);
-Product/Device (usability, functions, size, weight, language, symbols, aesthetic
characteristics, usefulness, reputation, adaptivity, mobility);
User experience Mobile web
browsing [31]
-Physical context (temperature, light, rain or humidity, objects the user is in contact with,
visible objects, crowdedness); -Social context (people, gender); -Temporal context
(period-time interval);
-Task context (user’s tasks);
-Body space (mental representation of body’s shape and postures through body’s junctions
(head, arm, hand, chest, back, leg, and foot) and senses: eyes, ears, mouth, etc.); -Space
around the body (is referred to through head/feet, front/back, and left/right mental
directional axes);
-Navigation space (mentally represented by landmarks and paths, links and nodes);
Spatial cognition Any activity
-Vista space (residential (home) or activity place (school, office));
-Local Displacement-Reinforcement Space (at a scale of district);
-Enlarged Displacement-Reinforcement Space (at a scale of region around the Local
collaboration [27]
-User/Company (User expression [company name, registration number, contact
information, address], Profile, Industry Sector, Product Service [description, category,
availability, specification, lead time, Pricing information, etc.]); -Temporal context (Time
Expression [start point, end point, granularity], Periodicity [frequency, granularity], Lead
time [Manufacturing time]);
-Location (Geographical Expression [postal address, GPS, country, region], Transportation
Mode [type, cost, frequency], Political Stability Index, Tariffs [Type, description, mode,
min/max limit]);
e-Commerce [28] -User context (Personal context, Task context, Environmental context, Social context);
-Business context (Product Context [product category, price, features], Business rules).
3. Connecting Context-Awareness and
Situation-Awareness to Enhance Decision
Making in Mobile Environments
From an external empirical observation and analysis, mo-
bility is a physical movement in space and time not only
going far from a point, but also getting closer to another
one, and involving a change of locations and surroundings
during which, the mobile person can see, hear, smell or
feel different elements that could affect his activities and
his state of mind. Three essential features can be distin-
guished as being part of a mobile environment:
The user’s personal bubble or personal context, in
which the decision maker acts, thinks, and tries to
decode the information he/she perceives. This per-
sonal bubble can be linked to the mental body space
developed by [39].
The surrounding environment in which external
things change while the decision maker is in motion.
This area may vary from closest to farthest areas sur-
rounding the user.
A set of natural (eyes, ears, nose, gestures, touch, etc.)
or artificial (GUI, mouse, keyboard, sensors, etc.) in-
terfaces which allow an exchange of information be-
tween the decision maker and his surroundings.
This empirical structure of mobile environment we
determine can be, somehow, mapped with the mental
spaces identified by [39]. Indeed, as reviewed in the “re-
lated works” section, [39] states that, from an internal
perception, human beings divide mobile environments
into three hierarchical spaces which are: 1) the body
space; 2) the space around the body; and 3) the naviga-
tion space. Therefore, the external personal bubble may
be linked to the mental body space, while the space
around the body and the navigation space can be consid-
ered as the mental representation of the surrounding en-
vironment. This mapping might be also made with the
vista space and the local displacement space proposed by
[40], but only if the current environment is the decision
maker’s life space.
The set of natural interfaces (human senses) identified
above are those that acquire events occurring in the sur-
roundings and provide to the user, his/her raw situation
awareness, while the set of artificial interfaces may be
exploited by context-aware applications to acquire, filter
and process these events in order to enhance the user’s
situation awareness and lead him to well informed deci-
Indeed, context-awareness, by enabling applications to
capture and process contextual information, can help
mobile decision makers to collect, exploit and build their
decisions on more accurate and precise information about
their business context than they would expect since some
contextual information might have escaped human atten-
tion or ask for more cognition and effort. For example,
by being informed almost in real-time by context-aware-
services of available services within a given area of in-
terest (AOI) and within a planned time interval, and by
getting alerted about important youth events within this
AOI from social networks and web sensors, and by gain-
ing summarized and substantial data about youth habits
and consumptions during these events, a mobile sales-
Copyright © 2012 SciRes. JGIS
man could plan effectively his sales-journey and would
more likely sell more. Simply put, context awareness
helps minimize user interactions and make applications
more intuitive and intelligent in order to improve their
usability in different contexts. Thereby context aware-
ness can support decision makers to get well informed of
their context, be aware to their business situation and
lead them to informed decisions.
Besides, it is worth to highlight that if most work on
contextual information awareness mainly considers the
problem of applications’ context-awareness, it surprise-
ingly seems to ignore the fact that the user is also aware
of his context, being supported or not by a context-aware
application. We argue that the user sensibility to the con-
text (known as situation awareness) should also be con-
In fact, while context-awareness is the ability of ap-
plications to sense some contextual information (e.g. the
user position, the surrounding temperature, etc.), situa-
tion awareness is the cognitive process by which the de-
cision maker learns, understands and builds a representa-
tion of his current decision situation (context + problem
to solve). Several authors such as [46] and [47] reminded
that the concept of situation awareness comes from the
air force and it relates to the pilot’s knowledge about the
aircraft and his flying environment. In short, “situation
awareness is about knowing what is going on around the
decision maker” and “richer Situation Awareness is more
likely to lead to good decisions and then to good per-
formance” [48,49].
Accordingly, context-aware applications could lead to
better decisions and better performance if they are able to
improve the raw situation awareness of the decision
maker by sensing, collecting, filtering, processing, and
supplying relevant contextual information in accordance
with the user’s requirements and goals, especially in a
mobile context where workspaces are dynamic and
change so often. This connection between applications’
context awareness and the decision maker’s situation
awareness and the improvement that can stems from it is
illustrated in Figure 1. This diagram shows that when the
contextual information is intelligibly processed and pre-
sented on a mobile device, it can help the decision maker
to be more aware of his contextual situation and drive
him to well-informed decisions.
Furthermore, the user cognition (the way he thinks and
feels) should be provided to the context-aware applica-
tion to make it more adapted or helpful to him. For ex-
ample, by knowing that the user is agoraphobic or claus-
trophobic, an application could alert him about unfamil-
iar public/places to go far from or narrow spaces (e.g.
elevator) to avoid.
Moreover, since mood (e.g. optimistic, pessimistic,
trustful, suspicious, etc.) may impact business perform-
ance [50,51], a mobile context-based application, by
capturing the user’s mood of the day (e.g. anger, impa-
tience, etc.) via mood aware applications and devices—
e.g. “microphones, cameras, heart and body monitors
[52], etc.—might propose to the user, routes which avoid
rush places, suggest him to postpone crucial meetings,
draw his attention on decisions he takes while in bad
mood, etc. A foreman may also use such mood-awareness
analysis to assign the right employee to the right task for
he right client to serve. t
Figure 1. Simplified model of connection between the applications’ context awareness and the decision makers’ situation
Copyright © 2012 SciRes. JGIS
We think that a cognition-compliant model of context
should then be envisage and should match the hierarchi-
cal organization of mobile environments. For this reason,
the notions of space and time from mobility, user’s bub-
ble and surrounding environment from mobile environ-
ments, and hierarchical cognitive spaces from human
perception of space will be exploited to organize, identify,
discuss and structure relevant contextual information and
dimensions for the mobile GeoBI context model we pro-
pose in next section.
4. Relevant Contextual Information and
Suitable Model for a Mobile GeoBI
When dealing with mobile context, time and location
(space) appear as the primary notions that should be ac-
counted (cf. major dimensions in the literature—Table 1)
since mobility and mobile environments are obviously
time-dependent and location-based (see Section 3). Stated
in the literature as context dimensions designated by
time-context (or temporal context) and location-context
(or spatial context), they are usually addressed as com-
mon contextual information. To our opinion, these no-
tions involve more specific contextual impacts that de-
serve further discussions: 1) time should not be consid-
ered only as a context but also as a variable dimension
which potentially affects all other contextual information
and context dimensions; 2) since Mobile GeoBI context
is location-based, it should be considered as related to the
location the mobile business professional is located in,
namely the local context (e.g. Quebec city business con-
text), or as related to a remote location, namely a remote
context (e.g. Wall Street) whose business activities may
influence the local context. The way these local and re-
mote contexts can share information has also to be han-
dled through a suitable and optimized mobile GeoBI
context model we propose.
To ease its comprehension and underline its multi-
level hierarchical structure, we will expose and build it
step by step in this part of the paper. We will thus expose
step by step: 1) the relevance and specificity of time; 2)
the relevance of location and the problematic of context
sharing between a local GeoBI context and remote con-
texts; 3) the different levels of a mobile GeoBI context;
4) the major dimensions of a mobile GeoBI context; and
5) the key elements composing each dimension of a mo-
bile GeoBI context.
4.1. Relevance and Specificity of Time
Dimension in Context
Several authors have identified the temporal context (also
called time context) as a relevant contextual key element
but they do not agree when trying to explain the concepts
it refers to. Indeed, while [20] and [22] stated that time
context is about the time of a day, the week, the month or
the season of year, etc., others researchers such as [53],
[42], and [34], only consider time (date-hour, the current
time or the system clock time) as being the relevant item
of a temporal context. Contrary to all other authors, [31]
argued that “time alone does not directly affect the user
experience. If it is a winter night, it is likely to affect
temperature, lighting, social context, users mental re-
sources and needs, but not directly user experience. The
relevant contextual information is not in the time itself,
but in the attributes the time affects”.
We consider like [31], that time-dimension may poten-
tially affect all other contexts and dimensions elements
so that it has to be handled as a specific dimension dif-
ferent from other contexts. Moreover, as stated by [53],
retrieving time itself (e.g. date-hour-second) and periods
(starting time—ending time) is also necessary and suffi-
cient for knowing the time or the period during which
events took place in the context, in which order, etc. In
addition, with a given date, it is possible to determine the
corresponding weekday, month, season, etc. Time is also
indispensable not only for identifying repetitive phe-
nomena or for predicting future events, etc., but also for
context versioning (chronological status of the same
context) and for context historization (archiving ancient
contexts into non-modifiable state for later use).
Both context versioning and context historization can
be defined based on some attributes or criteria. For ex-
ample, Figure 2(a) presents an activity-based context
versioning and shows how the same context may evolve
over the time; Figure 2(b) provides a case of context
historization based on the visited locations so that ancient
contexts of the same location might be recovered and
compared more easily.
To highlight the starting and ending time on contextual
elements affected by time, the temporal dimension is
represented by a temporal pictogram (x) introduced by
[54] for indicating time-dependent attributes. A mobile
GeoBI context itself is time-dependent and all its dimen-
sions (see Table 2) are potentially time-dependent as
illustrated below by the UML model proposed in Figures
In addition to being time-dependent, a mobile GeoBI
context is also location-based and may share contextual
information with a remote context located in another
location. This point is discussed in the next section.
4.2. Relevance of Location-Dimension and
Problematic of Context Sharing
Location is naturally a relevant contextual dimension as
mobility in general and professional mobility in particu-
lar, is usually justified by the user’s need to experiment a
Copyright © 2012 SciRes. JGIS
(a) (b)
Figure 2. (a) Example of context versioning; (b) Example of context historization.
Table 2. Relevant dimensions for a mobile GeoBI c onte xt.
Relevance for
Context Level Dimen-sions Description Mobility BI
Goal The mobile person’s goal, agenda, needs,
intentions or interests in what he is doing or
intend to do.
Pertinent to assist the user to reach the expected task. For
example, detecting the user’s intention to visit the closest
client by sensing his current position and direction, and
then propose him the most rapid route and display the
most up to date and critical indicators about this client.
Identity The person’s role and identity such as his
civil, professional or use profile.
Moderately relevant for
accessing spatial
navigation support
Strongly relevant for
supplying right information
to the right decision maker
based on his identity.
The way the person thinks, acts or feels. In
short, the person’s psychological profile
(mood, behavior) and preferences (like/
Relevant to know for
example where the user
should not go (e.g.
agoraphobic or
claustrophobic), or
would like to visit.
Relevant for handling in
which ways the mobile
worker might be “advised”
and assisted in his
Personal context
Tasks carried out. These can be Mobility
tasks (e.g. driving, walking, etc.), BI tasks
(requesting decisional data, meetings),
Communication tasks (calling, messaging,
etc.), Other tasks (e.g. carrying a weight,
painting a wall, etc.).
Relevant for Mobility
Relevant for BI tasks.
All information about business strategy,
activities, resources, markets, competition
and partnership. In short, all about business
facts and problems: metrics, indicators,
KPIs, etc.
Not really relevant for
Of course, highly relevant
for business intelligence
Technological capabilities in the
surrounding environment (Hardware
(networks, devices, etc.), Software,
Data, possible interactions (HCI)).
Relevant for way
finding support or
context-aware services.
Relevant for accessing and
processing context-based
business data.
Surrounding context
(including ambient
Social context
Social context is not only about social
networks, but also about culture, power
systems (i.e. politics) and resources
management (economy). In short, it’s
about social organizations of humans
and resources.
Moderately relevant for
asking location
information or
discovering POIs of a
Strongly pertinent for
understanding and dealing
with social groups, local
culture, resources and
Copyright © 2012 SciRes. JGIS
Refers to environmental conditions (seasons,
weather, noise) and services (transportation,
banking, hotels booking, etc.) available in
this environment.
Strongly relevant for adapting activities (Mobility and
BI tasks) to environmental conditions and available
Surrounding context
(including ambient
Spatial context
Refers to spatial localization of pertinent
objects located in the mobile environment
including persons, natural geography objects
(e.g. lands, vegetation, water, natural
resources, etc.) and human geography objects
(roads, places, POIs, infrastructures, etc.).
Strongly relevant for
spatial navigation.
Strongly relevant for
locating companies,
customers and for delivery
issues (costs, delays, etc.).
All contexts and
dimensions above Temporal
Refers to the time or the period during which
tasks are carried out, events occur, resources
are available, etc. For us, the temporal
dimension is specific and affects all context
and dimensions we treated above. This
specificity is explained in Section 5.4.
Relevant for journey
duration, transportation
means availability time,
Pertinent for monitoring
business evolution over
time, timeliness, etc.
Figure 3. Optimized context-sharing model.
Figure 4. Modeling mobile GeoBI context levels in accordance with context-sharing.
Copyright © 2012 SciRes. JGIS
Copyright © 2012 SciRes. JGIS
Figure 5. Multilevel mobile GeoBI context model (GeoMoBICoMod) for context-sharing and structuring (top level model).
physical proximity to resources [17] such as meeting
persons (customers, suppliers, etc.), scrutinizing more
closely a phenomenon (pollution, traffic, sales on the
field, etc.) or visiting POIs (Points Of Interest), etc. in a
given location which may affect the user, ease or com-
plicate his business activities, etc.
In the scope of mobile GeoBI context, this location
dimension will be designated in the rest of the paper as
“(geo) spatial con text”, and will refer to the spatial local-
ization of relevant features located in the mobile en-
vironment including persons, natural geography features
(e.g. lands, vegetation, water, natural resources, etc.) and
human geography objects (roads, places, POIs, infra-
structures, etc.). This (geo) spatial context is strongly
relevant for supporting the decision maker’s spatial
navigation (mobility support) as well as for locating
companies, customers, delivery issues (costs, delays,
etc.), etc. for BI support. To specify that these objects are
geo-localized, we will use spatial pictograms (H,w,e)
introduced by [54] to indicate their geometry (point, line,
and polygon) and by the way, their localization coordi-
By referring to the mobile environment around the de-
cision maker’s position, a mobile GeoBI context is by
definition a local (spatially limited) context dealing with
contextual information present in this local context of the
mobile business man. As nobody is completely cut off
from the world, especially in nowadays global world, a
local business contextual information (e.g. gas price in
Quebec City) may be affected by a remote context what-
ever the business context (e.g. wall street stock exchange)
or any other context (e.g. war in Libya) is. So, a mobile
GeoBI context is about a local context potentially sharing
information with remote contexts.
This problematic of context sharing may be handled
through different ways resulting in location-based con-
textualization of contextual information to better handle
influences and information-sharing between local and
remote contexts.
A first approach could consist in considering that local
context as well as remote contexts are parts of a final
context, and that this final context should result from the
combination (composition operations) of the local con-
text and remote contexts. This is the approach adopted by
[34] and illustrated in Figure 6.
The limits of this approach reside in the fact that if we
know well the structure and content of the local context,
we do not really always have a wide knowledge of the
remote context. The remote context could be very similar
to the local context (e.g. GeoBI context), or totally dif-
ferent (e.g. War context) and its content may not be ac-
cessible at all except for the shared contextual informa-
The second approach we propose to optimize and pal-
liate the limits of the first one, considers that the local
business context and remote contexts are separate con-
texts which interact and share contextual information
with each other. This approach provides a flexible solu-
tion to handle shared information and focuses mainly on
shared information rather than on the entire remote con-
text which is not always well known. This approach is
depicted by the UML model illustrated in Figure 3. It is
an adaptation from [34]. Some classes have been pro-
vided with comprehensive but non-formal attributes just
to help in integrating the examples of local and remote
contexts given in the previous paragraphs; and for sim-
plification sake, we chose to add an exclusive constraint
to underline that a given remote context cannot encapsu-
late a GeoBI context type with a different context type.
This model is the first level (the sharing level) of the
multi-level mobile GeoBI model we are building and is
intended to map human multi-level perception of mobile
environments. As designed in the model (Figure 3), the
local context as well as the remote context can be mobile
GeoBI contexts. The polygon pictogram (e) has been
added in the left side of local and remote contexts to de-
fine them as limited and geo-located. Relevant contextual
information composing these contexts, in addition to
time and location, will be identified and organized in
Sections 4.3 and 4.4 in a way that match the hierarchical
structure of mobile environment and human perception
of space presented in Section 3.
4.3. Relevant Hierarchical Levels of a Mobile
GeoBI Context
Integrating the hierarchical cognitive spaces (body space,
Figure 6. Combination of local and remote contexts to the
same context by [34].
space around the body, navigation space) and main ele-
ments distinguishable in the environment (the user’s
bubble, the surrounding environment, the set of natural
and artificial interfaces) identified from internal and ex-
ternal scrutiny of mobile environments (Section 3), we
propose three hierarchical levels of context composing a
mobile GeoBI context as follows:
The personal context which is about any contextual
information in relation with the user such as his pro-
files, goals, tasks, etc. A fair number of mobile appli-
cations could be classified as personal context-based
applications given that they are only aware of the
user’s profile (username, password, access rights, etc.).
The ambient context that is related to the immediate
environment around the user in which things can be
more accurately sensed by the user (via human inter-
faces as eyes, ears, skin, nose, etc.) as well as by ap-
plications (via computer interfaces such as sensors,
e.g. thermometer, GPS, etc.). The ambient context is a
subset of the surrounding context.
The surrounding context which potentially covers the
whole surrounding environment. Its limits may vary
from a district to a city or country perimeter. More
generally, a possibility should be offered to the user
to spatially draw the borders of this context. For ex-
ample, by drawing his area of interest (AOI) around
his position (or not), the user could be able to request
or receive business information related to this area
(and outside it). An example of such a request may be:
“Which are the fast selling products versus the best
profitable products per store in a radius of 5km
around my current geographic position during the last
two months?”
Organizing the mobile GeoBI context into these hier-
archical levels of context is useful and helpful as it is
consistent with existing hierarchical systems usually in-
volved in mobility such as wireless IT networks which
spans from personal area networks (PAN) to world area
networks (WAN), mobile mapping (zooming levels),
mobile cognition and mobility levels. Figure 7 illustrates
this parallel among mobility, spatial cognition, mobile
mapping, and network technologies.
This organization of mobile context also helps han-
dling the fact that in the context of business intelligence,
a local (mobile) context may be affected by, interact or
share contextual information with a remote context as
previously presented. Rather than melting all contextual
information in the same and unique level while it is not,
this hierarchical structure conceptually provides a right
way to better organize contextual information by putting
the right information at the right context level. For the
mobile context-aware application and for the mobile user,
it will be also a straightforward approach (without addi-
tional computing time) to know if it or he/she is dealing
Copyright © 2012 SciRes. JGIS
Copyright © 2012 SciRes. JGIS
Figure 7. Hierarchical le vels (of granular ity) of mobile context in parallel with other standpoints.
with local or remote information, and if that information
is related to a mobile person, the immediate environment
or the distant surrounding environment.
Figure 4 presents how these hierarchical mobile
GeoBI context levels can be modeled and integrated with
the first level (context-sharing level) of the multi-levels
mobile GeoBI context model we are constructing section
after section We have chosen to geo-localize the personal
context as a point which represents the current position
of the concerned person, but this may vary depending on
the targeted usage: e.g. representing the space occupied
by the person’s body (e.g.: shoulder width X chest
thickness) rather than his position.
Now, for each level of context (personal, ambient,
surrounding), relevant contextual information and di-
mensions have to be identified and analyzed in the pur-
pose of being sensed and/or managed by context-aware
mobile GeoBI applications.
4.4. Relevant Context Dimensions and Basic
Model for Mobile GeoBI Context
Based on Business Intelligence activity, mobile envi-
ronments nature and the existing work on contexts listed
previously, the following dimensions (Table 2) have
been considered as relevant for a GeoBI context-aware
application. They are potentially able to fully support a
mobile decision maker carrying out tasks in a mobile
business intelligence context. To remain compatible with
human perception of space, these dimensions will lean on
the three hierarchical context previously identified: per-
sonal context, ambient context, and surrounding context.
Each context has been explored through its main
meaningful contextual data grouped into dimensions. We
consider that a dimension may be in turn a context en-
compassing other dimensions (e.g. technological context
is related to the technological dimension of the sur-
rounding context, and should be composed of other di-
mensions, such as data, devices, interactions or networks
context, etc.). The dimensions have been selected based
on their relevance for mobility and/or for Business Intel-
ligence activity.
By appending these context dimensions to each con-
text level modeled in the previous part of the model
(Figure 7) we are assembling since the beginning, we
complete here, our proposal of a multilevel mobile
GeoBI context model for context-sharing and structuring
(Figure 5). It depicts a hierarchical structure of the mo-
bile GeoBI model in which, context levels are rendered
into classes having an aggregation of dimensions. As
ambient and surrounding contexts have the same dimen-
sions, they aggregate the related dimensions classes by
inheriting an abstract class named “Context around the
Note that ambient and surrounding contexts share the
same dimensions. Their difference resides in their scope.
Indeed, ambient context is a restriction of surrounding
context to a given area of Interest (AOI), which leads to a
kind of projection (limitation) of the dimensions content
inside this AOI. Conversely, a surrounding context is
somehow, an extension of ambient context to a wider
AOI. Also note that neither the spatial dimension nor the
others have a spatial pictogram specifying the need to
geo-localize them because they will be mapped (i.e.
clipped) to the spatial extent of the context around the
user (ambient context and surrounding context). Then,
only detailed contextual information (e.g. devices—mo-
bile or not—for technological context, social groups for
social context, companies for business context, roads for
spatial context, etc.) belonging to these dimensions need
to be geo-localized.
4.5. Advantages and Innovations of the
Proposed Model
The proposed multi-level mobile GeoBI context model is a
first contribution of our ongoing research on context-based
mobile GeoBI. Its advantages reside in the fact that it:
Provides a right way to better organize contextual
information by putting the right information at the
right context level rather than melting all contextual
information in the same and unique level while it is
Matches human perception of space which typecast
contextual information into closest, near or remote
elements. This matching has the advantage to ease the
future presentation of processed contextual informa-
tion and make it appear as natural to the user;
Provides a flexible solution to handle information
shared between local and remote contexts and focuses
mainly on shared information rather than on the entire
remote context which is not always well known;
Is appropriate to the hierarchical data storing, group-
ing and aggregating usually implemented by BI sys-
tems. That could ease a future coupling of this model
with BI data structures in order to provide mobile us-
ers business professionals with appropriate contextual
The proposed model also provides some innovations
residing in the fact that it is explicitly:
Easy to extend by extending a given dimension or by
plotting à new one, from top to bottom,
And it adopts a location-based contextualization of
contextual information to better handle influences and
information-sharing between local and remote con-
This final version constitutes a top-level model and a
basic framework which could be extended with detailed
elements. Next section provides for researchers and prac-
titioners needing to extend this model, a wide and some-
how complete inventory of relevant detailed contextual
information organized by context level and context di-
4.6. Detailed Relevant Contextual Information
for a Mobile GeoBI Context
To be context-aware, mobile GeoBI applications will
need to sense (with sensors) or retrieve (e.g. through user
inputs) detailed precise and accurate information (e.g.
user position, his office address) that can be turned into
data usable by an IT system. Then, after determining and
modeling the main concepts involved with mobile GeoBI
context, we have inventoried (mainly based on the de-
tailed review conducted in Section 2.3—see Table 1)
some detailed contextual information which may help
extend, complete and feed these concepts with more de-
tailed classes and attributes. Identified elements have
been analyzed as relevant for business intelligence ac-
tivity in mobility and have then been organized by con-
text level and context dimension.
4.6.1. Relevant Contextual Information in
Personal Context
Personal context is about any relevant contextual infor-
mation regarding the user which might affect the behave-
ior of an application. Table 3 exposes some elements
which may help characterize a mobile GeoBI context.
Personal context is intended to bring answers to ques-
tions related to the mobile worker, namely:
“Who is he?” i.e. his identity (e.g. civil profile,
physiological profile, social profile, professional pro-
file, etc.);
“What does he need or intend to do?” i.e. his goals
(intentions, needs, interests, agenda…);
“How does he proceed to think of, to solve a problem/
How does he feel things, etc.?” i.e. his cognition (e.g.
psychological profile (mood, temperament, feelings),
preferences, etc.);
“What is he doing/what does he have to do?” i.e. his
tasks which can be grouped into mobility tasks (e.g.
walking, driving, etc.), communication tasks (e.g.
calling, messaging, tchating, twitting, etc.), BI tasks
(meetings, consulting/requesting decisional data),
physical work tasks (e.g. carrying a box of products),
4.6.2. Relevant Contextual Information in
Context around the User
Similarly to the personal context, the context around the
user (surrounding and ambient contexts) context has been
designed based on answers to some main questions re-
lated to contextual issues which usually are: when (tem-
poral context which is actually related to all levels of
context)? Where (spatial context)? In which environ-
mental conditions (environmental context) and in which
human environment (social context) events are happen-
Instead of what is usually stated [55], social context is
not only about Culture or Social network (social groups
and their relationships), but it is also about Power (au-
thorities) and Resources, and their distribution between
social groups and institutions. Social context brings then
an overview of the social organization of a given society.
As far as business intelligence is concerned, new con-
texts need to be introduced to take into account business
and technological aspects providing answers to the fol-
lowing questions: “what is concerned?” Business Intelli-
gence (Business Context); and “how to achieve that?” By
Copyright © 2012 SciRes. JGIS
Copyright © 2012 SciRes. JGIS
Table 3. Relevant contextual information for a mobile GeoBI context at the personal level.
Context level Context
dimensions Relevant contextual information
(Key elements and details)
Civil Profile (Name, Address, Marital Status, Languages, etc.).
Professional Pr o file (Diploma, Skills, Experience, Performance, Role, Results, etc.).
Physiological Profile (Body Description [Weight, Height, Skin Color, Hair color, Eye Color]; Body Disabilities
[motor, visual, hearing, etc.]; Genetics [DNA, Blood Group, Rhesus, Fingerprint, etc.]; Health [diseases, allergies,
Social Profile (Family, Community, Association, Friends, Relatives, Life Style, etc.).
Use(r) Profile (Login, Password, Rights, etc.).
Needs/Intentions (Business needs to fulfill or objectives to reach at personal level. E.g. Need to register 10 new
customers today!).
Interest (Business centers of interest. E.g. selling innovative products, prospecting customers, analyzing revenues
from delivery services, keeping informed of stocks level in the company’s closest shops, etc.).
Agenda (Planned interest or tasks. e.g. from 9 to 10 a.m. prospecting customers).
Psychological Profile (Temperament, Mood, Hobby, Personality Traits, etc.).
Cognition Preferences (Indicate what the decision maker likes the most for each of his profile. E.g. Civil Profile: A perfect
bilingual, but prefers speaking French, etc.).
BI Tasks (Type [consulting/requesting decisional data], metrics, Qty of Data to download, Network, costs, etc.)
Mobility Tasks (Type [walking, driving, etc], Duration, etc.).
Communication Tasks (Type [e.g. calling, messaging, tchating, twitting, etc.]; Duration, Network, Costs, etc.).
Physical Tasks (Duration, Difficulty, Distance, etc.) E.g. carrying a box of products, etc.
exploiting technological capabilities (Technological Con-
Based on these questions and answers, the following
key elements and their detailed information have been
identified as potential relevant contextual information for
a mobile GeoBI context in the ambient and surrounding
contexts levels (Table 4).
Thanks to these detailed contextual elements, the pro-
posed top level model of mobile GeoBI context (above
Figure 5) could be richly extended and detailed. Future
work will provide researchers and practitioners with a
more complete model for context-reasoning and imple-
mentation perspectives. Since this is a work in progress,
the implementation of the model will then be tackled
5. Conclusions and Future Work
The requirement for high performance results in the mo-
bile global highly competitive and technology-based
business world we are in, has led to the need for business
professionals to get supported by convenient mobile DSS,
in order to keep ruling and monitoring their companies
indicators from anywhere at any time.
This paper in its beginning has stated that such mobile
DSS are nowadays mostly built on BI Systems to allow a
full accounting and remote access to organizational data
and analysis tools, but suffer from do not taking into ac-
count the user’s mobile context. The need for con-
text-based and geospatial-enabled BI (GeoBI) applica-
tions has then been justified, and the problematic of iden-
tifying relevant contextual information to capture and to
model into a way that matches BI data models and hu-
man cognition of mobile spaces has been raised. Ele-
ments of solution have been provided though out the rest
of the article.
Indeed, after demonstrating that context-awareness
may lead to faster and better well-informed decision
making, a large inventory of relevant contextual informa-
tion identified and captured by authors in the literature
and intended to context-aware applications has been pro-
vided to apprehend the state of the art of existing work.
From this review, and based on BI activity and mobile
environments nature, we have proceeded step by step to
the identification and analysis of major relevant context
dimensions for mobile GeoBI contexts so that at the end,
block after block, a suitable multilevel Geospatial Mobile
Business Intelligence Context Model (GeoMoBICoMod)
has been proposed, designed and justified.
In short, the other major contributions of this paper
A highlight of the connection between context-
awareness, situation-awareness and decision making
in the process of decision-making;
A distinction of time-context as a specific dimension
affecting all other context elements;
Table 4. Relevant contextual information for a mobile GeoBI context in ambient and surrounding contexts.
Context level Context
dimensions Relevant contextual information (Key elements and details)
Company ([company name, registration number, contact information, address, etc.])
Strategy (Objectives/results to reach, Business models and plans, business management processes, problem
solving processes, analysis/forecasting tools and techniques [metrics, KPIs, dashboards, reports, etc.], etc.)
Activity (Nature, Management chain [managers and teams hierarchy], Production process [rate, input,
output, production time, etc.], Tasks [nature, objective, constraints, etc.], Involved resources [human
resources, goods/services, financial resources, etc.], etc.)
Business Resources (turnover, cash flow, debts, goods/services, products, human resources [leaders,
managers, workers, consultants, etc.], material resources [buildings, offices, machines, computing systems,
etc.] etc.)
Markets (Sectors, Supply and demand, Consumers, Customers, Suppliers, Investors, Distribution networks,
Stock exchange, etc.)
Competition / Partnership (Nature, rules, actors, etc.)
Hardware (computers, mobile phones, networks [wired networks, mobile networks, GNSS networks, etc.],
sensors [Thermometer, Accelerometer, GPS receiver, Gyroscope, scanners, etc.], etc.)
Software (Operating systems, System oriented software [libraries, system services, web services, etc.], User
oriented software [word processing [Ms word, spreadsheets, etc.], web services [google maps, facebook,
twitter, etc.], e-mail software [thunderbird, outlook], utilities [notepad, agenda, etc.], etc.])
Data (data sources [files, databases, data warehouses, etc.] , access and privacy, data model, metadata, data
integrity, quality and security, etc.)
Human-Computer Interactions—How to interact with hardware, software and data(Interfaces [Human
interfaces, e.g. nose, eye, hand, etc.; Computer interfaces, i.e. I/O peripherals e.g. keyboard, GUI, etc.],
Interactions [input interaction, output interaction, interaction modality (e.g. multimodality), etc.])
Social Group (Type [Family, Friends, Age Group, Community, Association], Needs, Affluence. i.e. resource
possession, level of power, demography, etc.)
Culture (Language [oral, gestural, symbolic], Life Style [custom in dressing, feeding, entertainment, etc.],
Norms [e.g. rules about obligations, authorizations, prohibitions, social hierarchy/organization (e.g.
relationships between men and women), etc.], Beliefs [religions, taboos, etc.], etc.)
Social Resources (Type [goods/services (e.g. Food, water, education, health, employment, etc.)], availability,
accessibility, management entity, etc.)
Social context
Power (Power type [Political, Economical, religious/ideological, cultural, etc.], power system [Anarchy,
Theocracy, Monarchy, Democracy, etc.], Institutions, Laws/rules/procedures about business, etc.)
Atmospheric phenomena (Climate[seasons and their global characteristics such as average precipitation,
temperature, humidity, etc.], Weather [day-to-day, and even hour-to-hour detailed atmospheric measures such
as temperature, rain, sunshine, cloud cover, winds, heat waves, blizzards, fog, frost, flooding, light, pressure,
Surrounding ambiance (Indoor ambiance, Outdoor ambiance, strike ambiance, disturbing ambiance [noise,
traffic jam, etc.], appropriate ambiance [calmness, privateness, etc.], joy ambiance [e.g. The company index
gained 10 points], deception ambience, etc.)
Available services (Service type[transportation, healthcare, banking, car renting, hotels, etc.], service
localization, service availability time, service cost, etc.)
Location (Location Type [Country, State/Province, City, District, Place, etc.], Location Name, Location
address, Location Geometries (in 1D, 2D or 3D), Location POIs (Points of Interests), etc.)
Natural Geography Objects (Land Object [Mountain, Valley, Plain, etc.], Vegetation Object [prairie, forest,
Savannah, etc.], Water Object [Watercourse, sea, ocean, etc.], Soil Object [Sand, Clay, etc.], Subsoil Object
[Gold, Oil, etc.], etc.)
Ambient and
Spatial context
Human Geography Objects (Housing infrastructures [buildings, houses, offices, etc.], Transportation
infrastructures [Means (car, train, plane), Routes (roads, rails, airports, etc.)], Technological infrastructures
[refers to technological context], etc.)
Copyright © 2012 SciRes. JGIS
A raise of the importance of location-based contextu-
alization of contextual information to better handle
influences and information sharing between local and
remote contexts;
An organization of mobile GeoBI context into hier-
archical context levels (personal, ambient and sur-
rounding contexts) in accordance with human hierar-
chical perception of space;
An inventory of detailed relevant contextual informa-
tion for a mobile GeoBI context;
The mobile GeoBI context model we have proposed in
this paper is a generic one and has been intentionally
limited to top-level concepts and basic challenges to ease
its comprehension and show its capacity to support mo-
bility as well as BI aspects while respecting an accept-
able page limitation of this paper.
Ongoing work deals with the extension of this generic
and top-level model with 1) detailed and somehow pre-
cise contextual information, context-aware applications
should be aware of, and with 2) contextual metrics that
will be introduced and integrated to boost its capacity
to support contextual business analysis. The extension
will then provide a more complete and fully BI-oriented
model for implementation perspective. For context-
reasoning purpose, the extension will be made using on-
tology-based formalism.
Further work will later deal with the implementation
of the model and the challenge of coupling this complete
context model with BI data structures and models (data
warehouses, data cubes, etc.) in order to provide the
users with appropriate contextual analytics on which they
can base their decision process and take fully informed
[1] A. Cuzzocrea, F. Furfaro and D. Saccam, “Hand-Olap: A
System for Delivering Olap Services on Handheld De-
vices,” International Symposium on Autonomous Decen-
tralized Systems 2003, Pisa, 2003.
[2] A. Maniatis, “The Case for Mobile OLAP,” 1st Interna-
tional Workshop on Pervasive Information Management
(in Conjunction with EDBT’04), Heraklion, 2004.
[3] É. Dubé, T. Badard and Y. Bédard, “Service Web de
Constitution en Temps Réel de Mini-Cubes SOLAP Pour
Clients Mobiles,” Atelier SIG UbiquitaireSIG Mobiles,
CQFD-Géo/Sageo, Clermont-Ferrand, 2007.
[4], “Getting Information Where and
When You Need It,” 2008. in f
[5], “Cognos 8 Go! Mobile Extend Business Intel-
ligence Value by Accessing Information on Mobile De-
vices,” 2009.
[6] D. J. Power, “Decision Support Systems: Concepts and
Resources for Managers,” Greenwood Publishing Group,
Westport, 2002.
[7] M. Golfarelli, S. Rizzi and I. Cella, “Beyond Data Ware-
housing: What’s Next in Business Intelligence?” DOLAP
04 Proceedings, New York, 2004.
[8] D. Oard and G. Marchionini, “A conceptual Framework
for Text Filtering,” Technical Report EE-TR-96-25 CAR-
TR-830 CLIS-TR-9602 CS-TR-3643, University of Mary-
land, College Park, 1996.
[9] J. Herring, “Building a Business Intelligence System,”
The Journal of Business Strategy, Vol. 9, No. 3, 1988, pp.
4-9. doi:10.1108/eb039219
[10] H. Luhn, “A Business Intelligence System,” IBM Journal
of Research and Development, Vol. 2, No. 4, 1958, pp.
[11] S. Negash and G. Paul, “Business Intelligence,” Hand-
book on Decision Support Systems, 2 Edition, Springer,
Berlin, 2008, pp. 175-193.
[12] E. Turban, J. E. Aronson and T.-P. Liang, “Decision
Support Systems and Intelligent Systems,” Pearson Pren-
tice Hall, Upper Saddle River, 2005.
[13] B. Evelson and N. Norman, “Topic Overview: Business
Intelligence,” Forrester Research, 2008.
[14] M. Frolich and T. Ariyachandra, “Business Performance
Management: One Truth,” Information Systems Manage-
ment, Vol. 23, No. 1, 2006, pp. 41-48.
doi:10.1201/1078.1058053 0/457 1/91771.5
[15], “Business Intelligence (BI),” 2009.
[16] C. Franklin, “An Introduction to Geographic Information
Systems: Linking Maps to Databases,” Database, Vol. 15,
No. 2, 1992, pp. 13-21.
[17] J. Urry, “Mobility and Proximity,” Sociology, Vol. 36, No.
2, 2002, pp. 255-274.
[18] C. Bolchini, C. A. Curino, E. Quintarelli, L. Tanca and F.
A. Schreiber, “A Data-Oriented Survey of Context Mod-
els,” SIGMOD Record, Vol. 36, No. 4, 2007, pp. 19-26.
[19] P. Lombardi, V. Cantoni and B. Zavidovique, “Context in
Robotic Vision: Control for Real-Time Adaptation,” In-
ternational Conference on Informatics in Control, Auto-
mation and Robotics, 2004.
[20] G. Chen and D. Kotz, “A survey of Context-Aware Mo-
bile Computing Research (Tech. Rep. TR2000-381),”
Department of Computer Science, Dartmouth College,
Hanover, 2000.
[21] B. Shilit, N. Adams and R. Want, “Context Aware Com-
puting Applications,” 1st International Workshop on Mo-
bile Computing Systems and Applications, 1994.
[22] L. Sarjakoski and A.-M. Nivala, “Adaptation to Con-
Copyright © 2012 SciRes. JGIS
text—A Way to Improve the Usability of Mobile Maps,”
Map-Based Mobile Services, Theories, Methods and Im-
plementations, Springer, Berlin, Heidelber, New York,
2005, pp. 107-123.
[23] L. Barkhuus and A. Dey, “Is Context-Aware Computing
Taking Control away from the User? Three Levels of In-
teractivity Examined,” Ubi-Comp2003 Conference, Lec-
ture Notes in Computer Science 2864, 2003.
[24] A. K. Dey, “Providing Architectural Support for Building
Context-Aware Applications, Doctoral Dissertation,” Geor-
gia Institute of Technology, 2000.
[25] R. Want, A. Hopper, V. Falcao and J. Gibbons, “The
Active Badge Location System,” ACM Transactions on
Information Systems, Vol. 10, No. 1, 1992, pp. 91-102.
[26] M. Weiser, “Some Computer Science Issues in Ubiqui-
tous Computing,” Communications of the ACM, Vol. 36,
No. 7, 1993, pp. 75-84. doi:10.1145/159544.159617
[27] P. S. Tan, A. E. S. Goh and S. S. G. Lee, “A Context
Model for B2B Collaborations,” 2008 IEEE International
Conference on Services Computing, 2008.
[28] P. Kumar, S. Gopalan and V. Sridhar, “Context Enabled
Multi-CBR Based Recommendation Engine for Ecom-
merce,” International Conference on E-Business Engi-
neering, Beijing, 2005.
[29] W. Zheng and Y. Yuan, “Identifying the Differences
between Stationary Office Support and Mobile Work
Support: A Conceptual Framework,” International Jour-
nal of Mobile Communications, Vol. 5, No. 1, 2007, pp.
107-122. doi:10.1504/IJMC.2007.011492
[30] H. Wigelius and H. VÄaÄatÄajÄa, “Dimensions of Con-
text Affecting User Experience in Mobile Work,” IN-
TERACT’09: Proceedings of the 12th IFIPTC 13 Interna-
tional Conference on Human-Computer Interaction, Ber-
lin, 2009.
[31] V. Roto, “Web Browsing on Mobile Phones—Character-
istics of User Experience, Espoo, Finland,” Ph.D. Disser-
tation, Helsinki University of Technology, Helsinki, 2006.
[32] B. Schilit and M. Theimer, “Disseminating Active Map
Information to Mobile Hosts,” IEEE Network, Vol. 8, No.
5, 1994, pp. 22-32. doi:10.1109/65.313011
[33] M. F. Mokbel and J. J. Levandoski, “Toward Context and
Preference-Aware Location-Based Services,” Proceed-
ings of the Eighth ACM International Workshop on Data
Engineering for Wireless and Mobile Access, 2009.
[34] T. S. W. Hofer, M. Pichler, G. Leonhartsberger and J.
Altmann, “Context-Awareness on Mobile Devices—The
Hydrogen Approach,” Proceedings of the 36th Annual
Hawaii International Conference on System Sciences,
Hawaii, 2002.
[35] N. Bradley and M. Dunlop, “Toward a Multidisciplinary
Model of Context to Support Context-Aware Comput-
ing,” Human-Computer Interaction, Vol. 20, No. 4, 2005,
pp. 403-436. doi:10.1207/s15327051hci2004_2
[36] A. Dey, G. Abowd and D. Salber, “A Conceptual Frame-
work and a Toolkit for Supporting the Rapid Prototyping
of Context-Aware Applications,” Human-Computer In-
teraction, Vol. 16, No. 2-4, 2001, pp. 97-166.
[37] T. Winograd, “Architectures for Context,” Human-Com-
puter Interaction (HCI) Journal, Vol. 16, No. 2, 2001, pp.
[38] B. Shilit, N. Adams and R. Want, “Context Aware Com-
puting Applications,” 1st International Workshop on Mo-
bile Computing Systems and Applications, 1994.
[39] B. Tversky, “Structures of Mental Spaces: How People
Think About Space,” Environment and Behaviour, Vol.
35, No. 1, 2003, pp. 66-80.
[40] I. Reginster and G. Edwards, “The Concept and Imple-
mentation of Perceptual Regions As Hierarchical Spatial
Units for Evaluating Environmental Sensitivity,” Journal
of Urban and Regional Information Systems Association,
Vol. 13, No. 1, 2001, pp. 5-16.
[41] A. Schmidt, M. Beigl and H.-W. Gellersen, “There Is
More to Context than Location,” Computers and Graph-
ics, Vol. 23, No. 6, 1999, pp. 893-901.
[42] A. Kofod-Petersen and M. Mikalsen, “Context: Repre-
sentation and Reasoning—Representing and Reasoning
about Context in a Mobile Environment,” Revue dIntel-
ligence Artif icielle, Vol. 19, No. 3, 2005, pp. 479-498.
[43] J. Gwizdka, “What’s in the Context,” Computer Human
Interaction, 2000.
[44] G. Chen and D. Kotz, “A survey of Context-Aware Mo-
bile Computing Research,” Technical Report TR2000-
381, Dartmouth, 2000.
[45] L. Arhippainen and M. Tähti, “Empirical Evaluation of
User Experience in Two Adaptive Mobile Application
Prototypes,” Proceedings of the 2nd International Con-
ference on Mobile and Ubiquitous Multimedia, Norrkö-
ping, 2003.
[46] L. Niu, J. Lu and G. Zhang, “Cognition-Driven Decision
Processes,” Cognition-Driven Decision Support for Busi-
ness Intelligence, Springer, Berlin/Heidelberg, 2009, pp.
53-73. doi:10.1007/978-3-642-03208-0_5
[47] M. Endsley, “Towards a Theory of Situation Awareness
in Dynamic Systems,” Human Factors, Vol. 37, No. 1,
1995, pp. 32-64. doi:10.1518/001872095779049543
[48] L. Niu, J. Lu and G. Zhang, “Managerial Cognition,” in
Cognition-Driven Decision Support for Business Intelli-
gence, Springer, Berlin/Heidelberg, 2009, pp. 31-37.
[49] M. Stanners and H. French, “An Empirical Study of the
Relationship between Situation Awareness and Decision
Making,” Technical Report DSTO-TR-1687 of the De-
fence Science and Technology Organisation, 2005.
[50] P. Jordan, S. Lawrence and A. Troth, “The Impact of
Negative Mood on Team Performance,” Journal of Man-
agement & Organization, Vol. 12, No. 2, 2006, pp. 131-
Copyright © 2012 SciRes. JGIS
Copyright © 2012 SciRes. JGIS
[51] J. R. Nofsinger, “Social mood and Financial Economics,”
Journal of Behavioral Finance, Vol. 6, No. 3, 2005, pp.
144-160. doi:10.1207/s15427579jpfm0603_4
[52] N. Randewich, “Intel Guru: Future Phones Will Sense
Your Mood,” International Business Times, 15 Septem-
ber 2010.
[53] N. Bradley and M. Dunlop, “Toward a Multidisciplinary
Model of Context to Support Context-Aware Comput-
ing,” Human-Computer Interaction, Vol. 20, No. 4, 2005,
pp. 403-436. doi:10.1207/s15327051hci2004_2
[54] Y. Bédard, “Principles of Spatial Database Analysis and
Design,” In: GIS: Principles, Techniques, Applications &
Management, 2nd Edition, Chapter 29, Wiley, New York,
1999, pp. 413-424.
[55] S. Tamminen, A. Oulasvirta, K. Toiskallio and A. Kankainen,
“Understanding Mobile Contexts,” Personal and Ubiqui-
tous Computing, Vol. 8, No. 2, 2004, pp. 135-143.
[56] R. J. Thierauf, “Effective Business Inteligence Sys-
tems,” Quorom Books, Westport, 2001.