Anda di halaman 1dari 12

See discussions, stats, and author profiles for this publication at: https://www.researchgate.


Big Data, the Internet of Things, and the Revised Knowledge Pyramid

Article  in  ACM SIGMIS Database · November 2017

DOI: 10.1145/3158421.3158427


3 637

1 author:

Murray E. Jennex
San Diego State University


Some of the authors of this publication are also working on these related projects:

respecifying the Jennex-Olfman KM Success Model View project

All content following this page was uploaded by Murray E. Jennex on 17 January 2018.

The user has requested enhancement of the downloaded file.

Big Data, the The knowledge pyramid has been used for several

Internet of Things, years to illustrate the hierarchical relationships

between data, information, knowledge, and wisdom.
An earlier version of this paper presented a revised
and the Revised knowledge-KM pyramid that included processes such
as filtering and sense making, reversed the pyramid
Knowledge Pyramid by positing there was more knowledge than data, and
showed knowledge management as an extraction of
the pyramid. This paper expands the revised
knowledge pyramid to include the Internet of Things
Murray E. Jennex and Big Data. The result is a revision of the data
Fowler College of Business aspect of the knowledge pyramid. Previous thought
was of data as reflections of reality as recorded by
San Diego State University
sensors. Big Data and the Internet of Things expand
sensors and readings to create two layers of data.
The top layer of data is the traditional transaction /
operational data and the bottom layer of data is an
expanded set of data reflecting massive data sets
and sensors that are near mirrors of reality. The
result is a knowledge pyramid that appears as an

Keywords: Knowledge Pyramid; Big Data; Internet of

Things; Analytics; Knowledge Management

Jennex and Bartczak (2013) revised the knowledge
pyramid to incorporate learning, filtering, and
transformation processes and technologies and to
reflect their perspective that there is a difference
between the knowledge management (KM)
knowledge pyramid and the general knowledge
pyramid. This model reflected that KM is about
generating actionable intelligence and identified filters,
processes, and technologies to accomplish this.
While the revised model was considered satisfying
and useful for KM research, it did not consider Big
Data, analytics, and the Internet of Things. This
paper addresses these issues and integrates Big
Data, analytics, and the Internet of Things (IoT) into a
further revised knowledge pyramid.
Acknowledgments The importance and contribution of this paper is
based on the rising importance of Big Data, analytics,
This paper is a progressive work with early versions and IoT. LaValle et al. (2013) report that top
being presented at the Hawaii International performing organizations use analytics five times
Conference on System Sciences (HICSS) (2009), at more often than lower performing organizations and
the SIGPHIL meeting at the Americas Conference for clearly demonstrate the value of Big Data and
Information Systems (AMCIS) (2011), and published analytics. On the other hand, the IoT is a rapidly
in the International Journal of Knowledge approaching future with close to 5 billion devices
Management (2013). Comments and discussion expected to be connected by the end of 2015 and
from these meetings and with readers of the papers rising to 20 billion by 2020, meaning that the IoT is
have been used to expand and modify the model that promising to change the value proposition of the
is presented in this paper. Internet (Barnaghi et al., 2012; Pomerleau, 2015).

The DATA BASE for Advances in Information Systems 69 Volume 48, Number 4, November 2017
This paper recognizes that value from data comes The literature review conducted for this paper utilized
through both processes that transform data into Google Scholar and the AIS eLibrary as document
information, knowledge, and ultimately actionable repositories. Searches were conducted using the
intelligence or wisdom and the subsequent use of the search terms “knowledge pyramid,” “knowledge
generated actionable intelligence or wisdom in value- hierarchy,” “Big Data,” “knowledge,” and “Internet of
generating activities. Additionally, it is important for Things.” Search results were reviewed specifically
organizations to understand the sources and useful for applicability to the creation of knowledge and/or
transformations of the data they collect. This fully intelligence. Big Data and IoT sources were focused
revised knowledge pyramid provides insights for on sources that were published since and including
organizations into the relationships between the IoT, 2010.
Big Data, data, information, knowledge, and
actionable intelligence and the transformation and The Traditional Knowledge Pyramid
filtering processes.
Jennex and Bartczak (2013) presented an overview
This is a conceptual paper that hopes to promote on the traditional knowledge pyramid. This paper
discussion and insight by researchers into the nature summarizes this background for completeness.
of KM and its relationship to Big Data, the IoT, and Ackoff (1989) is given credit for the first academic
the overall processes of learning that transform data publication of the knowledge pyramid (see Figure 1).
into information, then knowledge, and finally The inference from the figure is that data begets
actionable intelligence. Once again, this paper uses information begets knowledge begets wisdom. An
definitions of terms important to KM, including additional inference is that there is more data than
definitions of data, information, knowledge, and information, more information than knowledge, and
intelligence that should be considered as working, more knowledge than wisdom. This model has been
consensus definitions, but it is recognized that these used in countless KM presentations and papers, and
terms are philosophical in nature and can be debated it is stated as a given truth that it is a generally
as long as we want. This debate is embraced, but not accepted model showing the DIKW hierarchy (Fricke,
encouraged, as I agree with Keen and Tan (2007) 2007; Hey, 2008; Sharma, 2004). The traditional
who believe that while it is important to understand knowledge pyramid uses the following summarized
KM terms, it is unproductive for researchers to get basic definitions:
focused on trying to precisely define these terms at
the expense of furthering KM research. The KM • Data – basic, discrete, objective facts about
discipline needs to allow the debate but also needs to something such as who, what, when, where.
unite into a consensus set of working definitions. It is • Information – data that is related to each other
hoped this paper will continue to spur this consensus. through a context such that it provides a useful
story, as an example, the linking of who, what,
Methodology when, where data to describe a specific person at
a specific time.
This is a conceptual paper; however, the arguments
made and conclusions presented are based on action • Knowledge – information that has been culturally
research in the form of system analysis as defined by understood such that it explains the how and the
Burstein and Gregor (1999). The inspiration for this why about something or provides insight and
paper comes from a project with a United States- understanding into something.
based defense contractor. Specifics of the project and
the company cannot be presented due to non- • Wisdom – placing knowledge into a framework or
disclosure agreements. What can be said is that the nomological net that allows the knowledge to be
applied to different and not necessarily intuitive
company attempted to take technologies and
experience developed / gained working with United situations.
States Department of Defense and other national These definitions imply that there is a relationship
intelligence agencies and generate a commercial between data, information, knowledge, and wisdom,
knowledge management offering focused on and it is the knowledge pyramid that provides a
providing KM tools and management processes. The graphical representation of these relationships as a
author participated in this project as a KM academic rollup hierarchy of data leading to information leading
expert responsible for providing KM focus and to knowledge leading to wisdom (henceforth, this will
direction. It is participatory research per Burstein and be referred to as the DIKW creation flow). Houston
Gregor (1999) as the author had a vested interest in and Harmon describe this in the use of summations:
the success of the project and in generating a
commercial KM offering and was able to reflect on I = ∑(D), K = ∑(I) = ∑∑(D), and W = ∑(K) = ∑∑(I)
the project while participating. = ∑∑∑(D) (Houston & Harmon, 2002).

The DATA BASE for Advances in Information Systems 70 Volume 48, Number 4, November 2017
to capture more data improves. A human example of
an improved sensor is using LASIK to improve
eyesight allowing a person to “see” much more,
enriching the vision that is captured. A mechanical
example is changing accuracy, range, and/or
sensitivity of sensors so they can “sense” more
The dashed lines with arrowheads reflect the
processing of data into information, knowledge, and
wisdom using the processes of insight, analysis,
analytics, and sense making. These dashed lines are
Figure 1. The Knowledge Pyramid (Ackoff, 1989) bi-directional indicating that the generation of
information, knowledge, and/or wisdom may progress
A counter point to the traditional knowledge pyramid up the hierarchy or produce feedback down the
is provided by Tuomi (2000). Tuomi (2000) states that hierarchy as the user understands more of what they
data is not the building block for information, are learning (recognizing Tuomi’s (2000) counter
knowledge, and wisdom as data is not observed, view). The lines between layers reflect the social
collected, or recorded in a vacuum. Rather, our networks used to transfer to different users. Social
understanding of the world through our wisdom and networks are being used loosely in this context and
knowledge drives us to collect specific information refer to any formal or informal, direct or indirect
and data to support our use of our knowledge and methods used to transfer and transform data,
wisdom. In this view, the hierarchy flows down the information, knowledge, and wisdom between users.
pyramid rather than up the pyramid, and data does Examples include classrooms, presentations, word of
not exist as a collection of unrelated facts as all mouth, published articles, webinars, email, the use of
collected facts are related to our basic knowledge Nonaka and Takeuchi’s (1995) Socialization,
and wisdom. While this is an improvement to the Externalization, Combination, and Integration (SECI)
knowledge pyramid, it still allows for confusion as it model of tacit knowledge transfer, and data mining
provides little insight and explanation for the analysts’ interpretations of patterns. Ultimately, this is
relationships between DIKW. The proposed revision a representation of the knowledge hierarchy and the
of the knowledge pyramid will show that flow is in general learning process for people, organizations,
both directions and introduces discussion on the and societies that results in multiple, large bodies of
relationships between DIKW. Finally, Dalal (2012) knowledge being generated and used. The endpoints
discussed the need for wisdom networks and (in both directions) lead to learning. Why learning?
research into the technologies and processes needed Learning has a multitude of definitions but for this
to produce wisdom from knowledge. In addition to model learning is defined as the acquisition of DIWK
their work in trying to produce a KM product, Tuomi that leads to a change in behavior or expectation
(2000) and Dalal (2012) provided the inspiration for within the individual or group that is doing the
the following revised knowledge pyramid. learning.
The knowledge pyramid is inverted, i.e. there is more
The Revised Knowledge Pyramid information than data, more knowledge than
The revised knowledge pyramid (Jennex & Bartczak information, and more wisdom than knowledge. One
(2013); see Figure 2) attempts to place the reason is simply mathematical. If information is the
knowledge hierarchy within the context of the natural structuring of data into meaningful combinations, then
or real world. What it shows is that data, information, the number of possible combinations for a quantity x
knowledge, and wisdom exist in a broader context, i.e. of data is minimally x! implying there is possibly a
humans are constantly gathering and processing greater amount of information than the original
data into information, knowledge, and wisdom. amount of data. Considering that users may have
However, the data gathered and processed is not all differing frames of reference for processing data in
that is available and is limited by the abilities of our different disciplines of thought (for example
sensors to detect, interpret, and capture data. accountants versus marketers or engineers versus
Sensors can be our human senses, others’ human biologists), it is very conceivable that the amount of
senses, or mechanical, where the sensors are information generated is greater than the original
anything that is not human such as a light detector, amount of data. This same argument can be used for
radio wave detector, pressure meter, a typed-in the generation of knowledge and wisdom, especially
transaction record, a query, etc. These sensors when it is also considered that users may have
reflect that as sensor technology improves, our ability different ethical, religious, or cultural belief systems

The DATA BASE for Advances in Information Systems 71 Volume 48, Number 4, November 2017
that could cause them to interpret information and key critical success factor. The use of intelligence
then knowledge differently (for example, Christians leads to organizational learning. Organizational
may generate different wisdom from the same learning is defined as a quantifiable improvement in
knowledge than Buddhists would or free societies activities, increased available knowledge for decision-
different than totalitarian societies). This is consistent making, or sustainable competitive advantage
with Nonaka (1994), Jennex (2008), and others that (Cavaleri, 1994; Dodgson, 1993; Easterby-Smith,
argue that all knowledge is context-specific. An 1997; Miller, 1996). Huber, Davenport, and King
example of this idea involves technology transfer that (1998) believe an organization learns if, through its
takes DIKW from one discipline (or context) and processing of DIKW, its potential behaviors are
attempts to apply it to another discipline (or context). changed.
Specifically, we can look at the case of taking the
Figure 2 combines the two pyramids. The implication
knowledge of using learning algorithms to identify
is that the knowledge-KM pyramid is a subset of the
meaningful sound patterns used to create sound
revised knowledge pyramid. There are other
canceling headphones and applying it to analyzing
differences. KM tends not to focus on wisdom but is
light diffraction patterns for identifying radioactive or
about beginning to apply intelligence concepts and
other (such as explosive) substances.
differentiates between knowledge needed to make a
The center of Figure 2 represents the knowledge-KM decision and specific actionable knowledge needed
pyramid. This pyramid is more like the traditional to make a specific decision in a specific context. For
pyramid, albeit the “wisdom” aspect is absent and example, marketing knowledge is needed to create
replaced with intelligence. The solid vertical arrows marketing campaigns but specific customer
represent application of KM processes (i.e. capture, knowledge is needed to make decisions as to how to
store, retrieve, apply) to the DIKW hierarchy pyramid. market to specific customers. For some users this is
Jennex (2005) defined KM as “the practice of confusing as it seems to imply there are differing
selectively applying knowledge from previous types of knowledge needed for KM. However, this
experiences of decision making to current and future tends to be consistent with Tuomi’s (2000) concept of
decision making activities with the express purpose wisdom.
of improving the organization’s effectiveness” (p. iv).
This confusion between wisdom and intelligence may
This implies that KM is not trying to address all
be a driver for those who practice Business
knowledge or wisdom. Rather, KM targets specific
Intelligence (BI) or Customer Intelligence (CI), etc. as
knowledge and wisdom needed by an organization to
they see a need for differentiating between general
perform specific tasks. Specific, actionable
information and knowledge and specific decision
knowledge and wisdom is defined as intelligence
information and knowledge within an organization. BI
(hence the replacement of wisdom by intelligence in
is generally defined as the collection, analysis, and
the pyramid), and it is the goal of KM to provide
presentation of business information for decision
intelligence to the organization for use in decision
making (Cody et al., 2002; Negash, 2004; Watson &
making. The lines between layers reflect the filters
Wixom, 2007) while CI and Competitive Intelligence
used to focus on specific DIKW and guide the
(also CI) are specialized forms of BI tailored to
application of KM processes such as capture, store,
specific intelligence (Cody et al., 2002; Negash,
retrieve, and apply. The purpose of KM is to make
2004). Figure 2 captures this idea by using the term
use of specific data, information, knowledge, and
intelligence rather than wisdom where intelligence
wisdom to generate intelligence. Additionally, KM
refers to very specific actionable knowledge. The
seeks to have data, information, knowledge, and
term intelligence is taken from the intelligence
intelligence shared among the right people at the
“activity” realm, which implies a function that seeks to
right time (Jennex et al., 2009). This implies that KM
generate actionable knowledge to be used in the
processes allow the filtering of data, information, and
formulation of strategies and tactics to accomplish a
knowledge to generate specific, actionable
specific goal, such as beating an opponent during
intelligence that is shared with specific, limited users.
wartime, or to determine what actions should be
Filters are placed on the social networks to limit
planned. This term is not chosen lightly as it does fit
access and to separate and capture that which is
into the DIKW hierarchy, meaning that intelligence is
needed from that which is not. In this vein, “filters” is
the interpretation of data, information, and knowledge
a fairly new term for KM. As such, it is the position of
to create courses of action or make specific decisions.
this paper that KM filters are the implementation of
Also, the solid arrows that represent KM are shown to
KM strategy. Given this link between filters and
be double-headed as the hierarchy may travel in
strategy, their importance is evidenced by Jennex
either direction. In fact, it is expected that while the
and Olfman (2005) who found KM strategy to be a
learning process will tend to generally be a bottom-up

The DATA BASE for Advances in Information Systems 72 Volume 48, Number 4, November 2017
Learning Organizational Learning Learning

Filters Insight,
Wisdom Intelligence Analysis,
Sense Making



Sensor Social Networks


Figure 2. The Revised Knowledge-KM Pyramid

process, meaning that it starts with the interpretation Additionally, filters support the suggestion of Dalal
of data, organizational learning and KM will generally (2012) with respect to identifying technologies to
be top-down processes where the actions or support wisdom/intelligence systems by helping KM
decisions that need to take place or be made are first focus on actionable intelligence and thus truly
determined and from there the intelligence, then becoming intelligence systems.
knowledge, then information, then data that is
A final difference between the traditional knowledge
needed to support taking the specific actions or
pyramid and the revised knowledge pyramid is the
making the specific decisions is identified.
removal of apices. This was done to remove
The other major difference is in the application of confusion as an apex tends to imply that there is an
filters. While the general learning process seeks to ultimate point, such as the ultimate key wisdom for
push data, information, knowledge, and wisdom out the traditional knowledge pyramid or a single key
to all who wish to use it, KM does not. As stated datum for the revised knowledge pyramid. While it is
earlier, KM seeks to support specific decision making somewhat satisfying to refer to ultimate points, this
and thus needs specific data, information, knowledge, paper does not support the idea of a “big bang”
and intelligence (the same can be said for BI). theory for data or an ultimate, supreme wisdom.
Additionally, KM seeks to share this with the right Rather, the paper does support that there is some
people at the right time. This implies that KM and BI initial level of data and some ultimate amount of
filters data, information, and knowledge to generate wisdom or intelligence. This position is supported
specific, actionable intelligence that is shared with when considered in light of the contextual nature of
specific, limited users. Moreover, these filters can knowledge and wisdom/intelligence and the multiple
also be described as analytical tools used to frames of reference and contexts that users bring to
segregate DIK into that which is needed from that the generation of knowledge and wisdom. Also
which is not. In this view, filters fit into the definition considered is that most organizations do not
of BI Systems: ultimately need one key piece of intelligence as there
is rarely a single decision that needs to be made.
BI systems combine data gathering, data storage,
Hence, the revised knowledge pyramid is pyramidal
and knowledge management with analytical tools
in form as that provides a visual impact as to the
to present complex internal and competitive
relative amounts of data, information, knowledge, and
information to planners and decision makers.
wisdom/ intelligence, but it is flat-topped, much like
(Cody et al., 2002, p. 178; a position further
the Aztec and Mayan pyramids and not like the
supported by Negash, 2004)
Egyptian pyramids (note that this is being used to

The DATA BASE for Advances in Information Systems 73 Volume 48, Number 4, November 2017
illustrate the point and not to imply that there is more expand on the IoT and discuss how the IoT
merit to one style of pyramid over another). generates Big Data. They then describe the
knowledge pyramid of the IoT as being raw data
The Revised Knowledge Pyramid leading to structured data with semantics leading to
Including Big Data, Analytics, and IoT abstractions and perceptions leading to actionable
intelligence (Barnaghi et al., 2012). Gubbi et al.
Wikipedia (2015) defines Big Data as a broad term (2013) and Atzori et al. (2010) expand on the IoT and
for data sets so large and complex that traditional see it as a vast sensor network with devices
data analysis methods are inadequate. McAfee and generating tremendous amounts of data by nearly
Brynjolfsson (2012) expand on this by observing that continuous recording of data reflecting the devices’
Big Data is created at a rate that humans cannot states and using ubiquitous analytics and cloud
analyze it without the aid of algorithms and analysis technology to generate value through networks of
tools. McAfee and Brynjolfsson (2012) further state devices.
that Big Data is being created at a rate of 2.5
exabytes per day and is mostly unstructured, which is The advent of Big Data and the IoT lends to a
the main difference between traditional data, which is revisualization of the revised knowledge KM pyramid.
structured, and Big Data which is not. Chen et al. Two issues need to be addressed. The first is the
(2012) agree with McAfee and Brynjolfsson (2012) definition of data:
but suggest analytics tools such as text analytics, Data – discrete facts such as who, what, when,
web analytics, network analytics, mobile analytics, and where that can be stored in a database
and data analytics are the key to transforming Big (Jennex and Bartczak, 2013)
Data into data, information, knowledge, and
intelligence. Madden (2012) also agrees and Big Data does not fit this definition; humans cannot
expands the definition offered by explaining that it analyze these larger data sets. Provost and Fawcett
means data that is too big, too fast, or too hard for (2013) say data science is about extracting
existing tools to process where: information and knowledge from data. Big Data uses
analytic tools to process it into humanly
• “Too big” means that organizations increasingly understandable data chunks. Humans interpret the
must deal with petabyte-scale collections of data understandable data chunks into information; then
that come from click streams, transaction once the information is understood as to how and
histories, sensors, and elsewhere. why it means what it does it becomes knowledge;
• “Too fast” means that not only is data big, but it finally, once the knowledge is fitted into a decision
must be processed quickly — for example, to context it becomes actionable intelligence. This is
perform fraud detection at a point of sale or supported by the application of data mining which is
determine which ad to show to a user on a one of the analytics tools used to make sense of
webpage. large data sets. Yu et al. (2014) explored second-
order data mining where humans determine meaning
• “Too hard” is a catchall for data that does not fit of the first order data mining outputs to produce
neatly into an existing processing tool or that actionable intelligence. They found that humans need
needs some kind of analysis that existing tools to be in the process and suggest that the habitual
can not readily provide. domains of the analysts influence data mining
interpretations. Another example is Barnaghi et al.
All agree the artifacts of knowledge are changing
(2012), who applied semantic operations such as
because of Big Data and look at the goals of Big Data
ontologies to Big Data to produce structured data.
as being to identify intelligence for evidence-based
The implication of both studies is that Big Data needs
decision making, transforming intuitive-based
to be transformed to data before it can be
decision making to evidence-based decision making,
transformed to information and so on, implying that
and pushing decision making to lower levels of the
Big Data is a distinct layer from data. A second
organization. However, Big Data does not remove
implication is that since IoT is a sensor net for Big
the need for human insight and vision. Finally,
Data, then IoT is a distinct layer from Big Data. A
Koronios et al. (2014) found that a key success factor
third implication is that since IoT does more than just
for Big Data is having a strategy for determining what
generate Big Data and there is more Big Data than
Big Data is needed to generate business value.
data, then IoT may be more than Big Data. A fourth
Wikipedia (2015) defines IoT as the network of implication is that since IoT is a sensor net, it is a
physical devices that connect to the web, usually distinct layer between reality and Big Data.
through a wireless connection, and communicate with
The second issue is the changing nature of wisdom
other physical devices for improving service of all
and knowledge. Chen et al. (2012), McAfee and
devices and creating value. Barnaghi et al. (2012)

The DATA BASE for Advances in Information Systems 74 Volume 48, Number 4, November 2017
Brynjolfsson (2012), and Madden (2012) all agree data, data, information, and knowledge not needed to
that automated tools and analytics are changing the support the generation of specific actionable
nature of knowledge and wisdom as they focus on knowledge.
producing actionable intelligence to support
The final addition is the double-headed arrows that
evidence-based decision making and automated
link the IoT and big data layers directly to the
decision making. This implies that the social
wisdom/intelligence layer. The arrows are double
networks previously used between the data,
headed to represent that there is automatically
information, knowledge, and wisdom layers need to
generated intelligence (per Chen et al. (2012),
be expanded to include analytics. However, these
McAfee and Brynjolfsson (2012), and Madden (2012))
analytics and the Big Data they help transform reflect
that is directly input into the wisdom/intelligence layer
the findings of Koronios et al. (2014) that strategy is
and that there is guidance as to what is needed from
required to guide the use of Big Data. This implies
the IoT and big data layers (per Koronios et al.
that organizations need to focus on their Big Data
(2014)). The purpose of these arrows is to reflect the
even if they are not using KM. Additionally,
trend of autonomous/automatic decision making.
Weinberger (2012) suggests that the availability of
Examples of this include intelligent drones, other self-
Internet-based digital media sources are changing
driving vehicles, or other automatic controllers that
the shape, evolution, and perception of knowledge,
are programmed to react to events should connection
resulting in the traditional pyramid of knowledge
to their human controllers be lost or they are put into
becoming a formless “network of knowledge.” This is
an automatic action mode or should their human
due to the ineffectiveness of filters normally used by
controllers not react to an emerging situation.
organizations to verify sources. The use of strategy to
guide acquisition of Big Data is an example of the
application of filters, albeit weak filters, allowing Big Discussion
Data to operate within the KM function in many What is the value of revising the knowledge pyramid?
different ways. Thus, Weinberger’s (2012) findings Ultimately Figure 3 showing the fully revised
support the inverted pyramid for the revised model. knowledge pyramid that includes KM, analytics, big
Figure 3 reflects the revised knowledge pyramid that data, IoT, and various intelligence initiatives
includes KM, Big Data, analytics, and IoT. The (business intelligence, customer intelligence) reflects
pyramid is now an hour glass. The top inverted the way technology innovation is radically addressing
pyramid is as before but the transformation societal and organizational needs for more data and
processes between layers now include analytics and more actionable intelligence. The drive for evidence-
weak filters. The bottom pyramid looks more like the based decision making rather than intuitive decision
traditional knowledge pyramid but now has layers of making is pushing the creation and adoption of
reality, IoT, and Big Data. The traditional pyramid technologies that gather and analyze observations of
shape reflects that reality still contains more data reality faster and in amounts greater than what
than we can record. However, the second layer, IoT, humans can process. The issues that are created
reflects that our ability to collect observations is are how to manage data and technologies; which
greatly increased and results in the third layer, big technologies to use; and what data, information,
data. The transformation processes between these knowledge, and actionable intelligence is needed for
layers reflect the use of automated analysis tools. decision making. This revised knowledge pyramid
provides value by helping with these three issues by
One major change is the renaming of the wisdom showing relationships between technologies and
layer to the wisdom/intelligence layer. This reflects decision support artifacts (Big Data, data, information,
the findings of Chen et al. (2012), McAfee and knowledge, intelligence) and by providing a model
Brynjolfsson (2012), and Madden (2012) of the that assists in creating a strategy for selecting and
changing nature of knowledge and wisdom to be managing technologies and decision support artifacts.
more focused on actionable intelligence.
Society and organizations manage by planning.
Another change is the redefining of the Resources and time are limited, so planning applies
transformation processes between data, information, thought before action, allowing for a systematic
knowledge, and wisdom/intelligence to include approach to problem solving. The output of planning
Weinberger’s (2012) weak filters used by humans to is a plan or strategy, a statement of how something
determine credibility while at the same time adding will be done. Society and organizations need to have
the more quantitative analytics processes for a strategy for managing the layers and technologies
identifying patterns. The strong filters shown in the in the revised knowledge pyramid. Jennex (2010,
KM pyramid reflect the application of strategy to 2012) investigated the components of a KM strategy,
define what is needed from each layer and are used found having a strategy improved organizational
to prevent overload of the decision process with big performance, and applied it to managing technology.

The DATA BASE for Advances in Information Systems 75 Volume 48, Number 4, November 2017
KM Pyramid

Learning Organizational Learning Learning

Actionable Strong
Wisdom/ Filters Insight,
Intelligence Analysis,
Sense Making

Knowledge Knowledge

Information Information

Data Data

Big Data Social Networks, analytics,

Big Data and weak filters
Machine Networks
IoT and IoT –and Automated Analysis
Other Sensors Tools and Systems
Other Sensors


Figure 3. The Revised Knowledge Pyramid with KM, Big Data and IoT

The basic components of a KM strategy can be • Identification of technologies to be used to

generalized and made useful for managing support capturing and processing Big Data, data,
knowledge pyramid activities including: information, and knowledge
• Identification of users of the knowledge pyramid • Generation of top management support
layers and transformation processes
• Establishment of metrics for Big Data, data,
• Identification of actionable intelligence needed to information, and knowledge use
support organizational/societal decision making
• Establishment of feedback and adjustment
• Identification of sources of Big Data, data, process on the effectiveness of actionable
information, and knowledge intelligence use
• Identification of Big Data, data, information, and The goal is a top-down strategy approach based on
knowledge to be captured the decisions to be made and identifying the
technologies and decision support components
• Identification of how captured Big Data, data,
needed. Additionally, the more focused the strategy
information, and knowledge is to be stored and
the stronger the filters that are created blocking Big
Data, data, information, and knowledge not needed
to support decision making.

The DATA BASE for Advances in Information Systems 76 Volume 48, Number 4, November 2017
Strategic approaches to managing the knowledge all these technologies can and must work together to
pyramid make sense on the organizational level in improve the efficiency and effectiveness of decision
support of organizational decision making, but do making processes. This pyramid model is more
they work for society? The knowledge society is an satisfying conceptually as it fits our perception of
approach that uses knowledge and intelligence to reality and what technologies can do to a better
create processes and products that better serve degree and clarifies some of the assumed process
citizens. Among knowledge society initiatives are KM, aspects of the knowledge pyramid. Is this the final
smart grids, electronic health, smart networks, model? Probably not, but it is a further step in the
electronic government, and crisis response and right direction and a starting point for more detailed
management. Many of these initiatives are being discussion of the knowledge pyramid. This revised
empowered by IoT, Big Data, and analytics as well as knowledge pyramid that includes the decision support
the rest of the knowledge pyramid layers and technologies and initiatives such as KM, Big Data, IoT,
transformation processes to perform tasks such as analytics, and business/customer intelligence
more efficient energy and water management, provides insight and guidance into how to create
bringing medical expertise to areas poorly served by comprehensive decision support strategies. The
doctors, and improving efficiency of transportation model shows that decision making is a process
and manufacturing systems. supported by many technologies and processes. It is
also hoped that this revised pyramid can be used by
The approach of the knowledge society leads to the
researchers to guide them in developing integrated
addition of the direct feed of intelligence from the IoT
decision solutions.
and Big Data layers to the wisdom/intelligence layer.
Figure 3 also shows that the bottleneck in the
knowledge pyramid is human comprehensible data References
represented by the hour glass shape with very large Ackoff, R. L. (1989). “From Data to Wisdom.” Journal
amounts of observations gathered and processed of Applied Systems Analysis, Vol. 16, No. 1: pp.
automatically on one end, and very large amounts of 3-9.
knowledge and wisdom/intelligence generated and Atzori, L., Iera, A., and Morabito, G. (2010). “The
used by decision makers on the other end. Internet of Things: A Survey.” Computer Networks,
Consequently, the IoT is required to assist Vol. 54, No. 15: pp. 2787-2805.
automatically controlled processes in utilities such as Barnaghi, P., Wang, W., Henson, C., and Taylor, K.
electricity, water, and communications in order to (2012). “Semantics for the Internet of Things:
react quickly to changes in operating parameters Early Progress and Back to the Future.”
(Gubi et al., 2013). The bottleneck in the data layer is International Journal on Semantic Web and
not acceptable for such systems if the direct feed Information Systems, Vol. 8, No. 1: pp. 1-21.
between IoT, Big Data, and wisdom/intelligence is to Bates, M. J. (2005). “Information and Knowledge: An
be implemented. Such a situation is undesirable for Evolutionary Framework for Information Science.”
many as they need to feel that humans are in control Information Research, Vol. 10, No. 4: paper 239.
and understand what is happening instead of Boyd, D. and Crawford, K. (2012). “Critical Questions
worrying about various concerns. Vlacheas et al. for Big Data: Provocations for a Cultural,
(2013) propose frameworks for this link to function Technological, and Scholarly Phenomenon.”
successfully while mitigating concerns. Still, while Information, Communication, & Society, Vol. 15,
the model shows the link, it is perhaps the most No. 5: pp. 662-679.
controversial addition to the revised knowledge Burstein, F. V. and Gregor, S. (1999). “The Systems
pyramid. Development or Engineering Approach to
Research in Information Systems: An Action
Conclusions Research Perspective.” Proceedings of 10th
Australasian Conference on Information Systems,
This further revised knowledge-KM pyramid is even
B. Hope and P. Yoong (Eds.), Victoria University
more complex and thus less satisfying as a visual
of Wellington, New Zealand, pp. 122-134.
model than the model from Jennex and Bartczak
Cavaleri, S. (1994). “Soft Systems Thinking: A Pre-
(2013). However, we are in a period of rapid
Condition for Organizational Learning.” Human
technological advancement resulting in dramatic
Systems Management, Vol. 13, No. 4: pp. 259-
impacts of societal and organizational decision
making. We are also in a period of declining
Chen, H., Chiang, R. H. L., and Storey, V. C. (2012).
resources resulting in proponents of each technology
“Business Intelligence and Analytics: From Big
competing with and promising more than the other
Data to Big Impact.” MIS Quarterly, Vol. 36, No.
technologies. This revised knowledge pyramid
4: pp. 1165-1188.
attempts to mitigate the competition by showing that

The DATA BASE for Advances in Information Systems 77 Volume 48, Number 4, November 2017
Cody, W. F., Kreulen, J. T., Krishna, V., and Spangler, Jennex, M. E. (2008). “The Impact of Culture and
W. S. (2002). “The Integration of Business Context on Knowledge Management,” in Current
Intelligence and Knowledge Management.” IBM Issues in Knowledge Management. M. E. Jennex
Systems Journal, Vol. 41, No. 4: pp. 697-713. (Ed.), IGI: pp. 6-11.
Dalal, N. (2012). "Wisdom Computing: Toward a Jennex, M. E. (2010). “Implementing Social Media in
Framework for Wisdom Research in Information Crisis Response Using Knowledge
Systems." AMCIS 2012 Proceedings, Paper 14. Management.” International Journal of
Available at Information Systems for Crisis Response and Management, Vol. 2, No. 4: pp. 20-32.
spectivesIS/14 Jennex, M. E. (2012). “Identifying the Components of
Davenport, T. H. and Prusak, L. (1998). Working a Knowledge Management Strategy.” 18th
Knowledge. Boston, MA: Harvard Business Americas Conference on Information Systems,
School Press. AMCIS18, August 2012.
Dodgson, M. (1993). “Organizational Learning: A Jennex, M. E. and Bartczak, S. E. (2013). “A
Review of Some Literatures.” Organization Revised Knowledge Pyramid.” International
Studies, Vol. 14, No. 3: pp. 375-394. Journal of Knowledge Management, Vol. 9, No. 3:
Easterby-Smith, M. (1997). “Disciplines of pp.19-30.
Organizational Learning: Contributions and Jennex, M. E. and Olfman, L. (2005). “Assessing
Critiques.” Human Relations, Vol. 50, No. 9: pp. Knowledge Management Success” International
1085-1113. Journal of Knowledge Management, Vol. 1, No. 2:
Foley, E. and Guillemette, M. G. (2010). “What is pp. 33-49.
Business Intelligence?” International Journal of Jennex, M. E., Smolnik, S., and Croasdell, D.T.
Business Intelligence Research, Vol. 1, No. 4: pp. (2009). “Towards a Consensus Knowledge
1-28. Management Success Definition.” VINE: The
Fricke, M. (2007). “The Knowledge Pyramid: A Journal of Information and Knowledge
Critique of the DIKW Hierarchy.” Journal of Management Systems, Vol. 39, No. 2: pp. 174-
Information Science, Vol. 35, No. 2: pp. 131-142. 188.
Gubbi, J., Buyya, R., Marusic, S., and Palaniswami, Keen, P. G. W. and Tan, M. (2007). “Knowledge
M. (2013). “Internet of Things (IoT): A Vision, Fusion: A Framework for Extending the Rigor and
Architectural Elements, and Future Directions.” Relevance of Knowledge Management.”
Future Generation Computer Systems, Vol. 29, International Journal of Knowledge Management,
No. 7: pp. 1645-1660. Vol. 3, No. 4: pp. 1-17.
Hey, J. (2004). The Data, Information, Knowledge, Koronios, A., Gao, J., and Selle, S. (2014). "Big Data
Wisdom Chain: The Metaphorical Link. Retrieved Project Success – A Meta Analysis." PACIS 2014
June 16, 2008, from: Proceedings, Paper 376. Available at:
Houston, R. D. and Harmon, E. G. (2002). “Re- LaValle, S., Lesser, E., Shockley, R., Hopkins, M. S.,
envisioning the information concept: systematic and Kruschwitz, N. (2013). “Big Data, Analytics
definitions,” in Emerging Frameworks and and the Path from Insights to Value.” MIT Sloan
Methods: Proceedings of the Fourth International Management Review, Vol. 52, No. 2: pp. 21-32.
Conference on Conceptions of Library and Madden, S. (2012). “From Databases to Big Data.”
Information Science (CoLIS4). H. Bruce, R. Fidel, IEEE Internet Computing, Vol. 16, No. 3: pp. 4-6.
P. Ingwersen, and P. Vakkari (Eds.), Greenwood Markus, B. (2005). Learning Pyramids. WSVA 1 –
Village, CO: Libraries Unlimited, pp. 305-308. Virtual Academy and the Surveying/
Huber, G. P., Davenport, T. H., and King, D. (1998). Geoinformatics Community –Virtual Academy I.
“Some Perspectives on Organizational Memory.” McAfee, A. and Brynjolfsson, E. (2012). “Big Data:
Unpublished Working Paper for the Task Force The Management Revolution.” Harvard Business
on Organizational Memory. F. Burstein, G. Huber, Review, Vol. 90, No. 10: pp. 60-68.
M. Mandviwalla, J. Morrison, and L. Olfman, Miller, D. (1996). “A Preliminary Typology of
(Eds.) Presented at the 31st Annual Hawaii Organizational Learning: Synthesizing the
International Conference on System Sciences Literature.” Journal of Management, Vol. 22, No.
(Hawaii). 3: pp. 485-505.
Jennex, M. E. (2005). “What is Knowledge Negash, S. (2004). "Business Intelligence." Comm. of
Management?” International Journal of the Association for Information Systems, Vol. 13,
Knowledge Management, Vol. 1, No. 4: pp. i-iv. Article 15: pp. 177-195. Available at:

The DATA BASE for Advances in Information Systems 78 Volume 48, Number 4, November 2017
Nonaka, I. (1994). “A Dynamic Theory of Wikipedia, (2015). Internet of Things. Retrieved April
Organizational Knowledge Creation.” 2, 2015, from
Organization Science, Vol. 5, No. 1: pp. 14-37.
Nonaka, I. and Takeuchi, H. (1995). The Knowledge- Yu, X., Shi, Y., Zhang, L., Nie, G., and Huang, A.
Creating Company – How Japanese Companies (2014). "Intelligent Knowledge Beyond Data
Create the Dynamics of Innovation. Oxford, U.K.: Mining: Influences of Habitual Domains."
Oxford University Press. Communications of the Association for
Pomerleau, M. (2015). “Cheat Sheet for the Internet Information Systems, Vol. 34, Article 53. Available
of Things.” GCN, Technology, Tools, and Tactics at:
for the Public Sector. Retrieved January 30, 2015,
from About the Author
Provost, F. and Fawcett, T. (2013). “Data Science & Murray E. Jennex is a Professor of Management
Its Relationship to Big Data & Data-Driven Information Systems in the Fowler College of
Decision Making.” Big Data, Vol. 1, No. 1: pp. Business at San Diego State University, editor in
51-59. chief of the International Journal of Knowledge
Rowley, J. (2007). “The Wisdom Hierarchy: Management, co-editor in chief of the International
Representations of the DIKW Hierarchy.” Journal Journal of Information Systems for Crisis Response
of Info. Science, Vol. 33 No. 2: pp. 163-180. and Management, and president of the Foundation
Sharma, N. (2004). The Origin of the “Data for Knowledge Management (LLC). Dr. Jennex
Information Knowledge Wisdom” Hierarchy. specializes in knowledge management, crisis
Accessed June 17, 2008: response, system analysis and design, IS security, e- commerce, and organizational effectiveness. Dr.
_origin.htm, 2008 Jennex serves as the Knowledge, Innovation, and
Tuomi, I. (2000). “Data is More Than Knowledge: Entrepreneurial Systems Track co-chair at the Hawaii
Implications of the Reversed Knowledge International Conference on System Sciences. He is
Hierarchy for Knowledge Management and the author of over 150 journal articles, book chapters,
Organizational Memory.” Journal of Management and conference proceedings on knowledge
Info. Systems, Vol. 16, No. 3: pp. 103-117. management, crisis response, end user computing,
Vlacheas, P., Giaffreda, R., Stavroulaki, V., Kelaidonis, international information systems, organizational
D., Foteinos, V., Poulios, G., & Moessner, K. memory systems, ecommerce, cyber security, and
(2013). “Enabling Smart Cities through a software outsourcing. Dr. Jennex is a former U.S.
Cognitive Management Framework for the Navy Nuclear Power Propulsion officer and holds a
Internet of Things.” Communications Magazine, B.A. in chemistry and physics from William Jewell
IEEE, Vol. 51, No. 6: pp. 102-111. College, an M.B.A. and an M.S. in software
Watson, H. J. and Wixom, B. H. (2007). “The Current engineering from National University, an M.S. in
State of Business Intelligence.” Computer, Vol. 40, telecommunications management and a Ph.D. in
No. 9: pp. 96-99. information systems from the Claremont Graduate
Weinberger, D. (2012). Too Big to Know: Rethinking University. Dr. Jennex is also a registered
Knowledge Now That the Facts Aren't the Facts, professional mechanical engineer in the state of
Experts Are Everywhere, & the Smartest Person California and a Certified Information Systems
in the Room Is the Room. New York, NY: Basic Security Professional (CISSP), a Certified Secure
Books. Software Lifecycle Professional (CSSLP), and a
Wikipedia, (2015). Big Data. Retrieved April 2, 2015, Project Management Professional (PMP).

The DATA BASE for Advances in Information Systems 79 Volume 48, Number 4, November 2017

View publication stats