Business Decision Analytics: Tacit Information and Leximancer Analysis

1
Business Decision Analytics
Executive summary
Unstructured information for a long time had been a point of avoidance by analysts.
However, with the advent of tacit information collection and algorithms to analyze textual
unstructured information such as that used in leximancer software, new data insights on non-
contemporary issues which include: cultural relations and their effects, skills held by staff
workers etcetera have been unfolded.
Research goals
In this project, tacit information for allocation of resources and identification of patterns
is examined using email exchange information mapped for 80 R, D &D staff workers. From the
information patterns, the data is divided into clusters with a specified node selection criterion i.e.
related nodes are grouped in one cluster. It is noted that unconnected nodes are not included in
the clusters due to their probable nature of outliers and hence insignificant in analysis.
From the second section on leximancer analysis, the results indicate there is more open
communication between the management and customers than there is between customers and
field service team. Additionally, it is noted that there is relatively sufficient support from the
field service team
Assumptions
In this project, the assumption that hanging but connected nodes indicate different
communication trends that is on different subjects is made.
Key words
Leximancer, Tacit knowledge, Concept map.

2
Business Decision Analytics
Introduction
Growth and sustainability of a business firm is often dependent on a wide range of
factors. From marketing strategies, research processes, technological advancements to adaptation
of shifting business structures, the business world is operating almost on entirety of a different
clock time a fact that is defining how new and already business firms are operating (Wawira,
2013). Over the past decade, technology advancement has grown to great measures bringing
new definitions of business processes.
Such business processes include: production, sales decision making etcetera. Shinnick
and Ryan (2008) on the role of information in decision making note that, “...the advent of the
World Wide Web and other communication technologies has significantly changed how we
access information, the amount of information available to us, and the cost of collecting that
information.” In practice, a good information system is definitive of the business operations as
well as it is key in decision making.
Unlike the tradition trial and error method of business decision making, modern decision
makers are faced with the challenges of making decision which are influential of the path a
business is likely to take. Whereas as good decision will oversee growth of a business, single
wrong decision, is likely to cost the business firm a whole lot of resources both monetary and
even worse might raise ethical concerns. However, availability of information to the decision-
makers has drastically improved the process of business decision making.
Tremendous technological advancements have influenced drastic changes in business
working environment. An article by the Forbes magazine (2016), argues that work environments
are outcomes of evolutionary processes and as such require managerial adaptation to ensure
smooth operations.

3
Business Decision Analytics
Body
The correct information, at the right time has got numerous advantages to business
organizations. Hence tacit information (Pivar, Horvat & Malbašić, 2012). Often, study of tacit
information can be tedious due to its unstructured nature unlike explicit information which is
accurate and straight forward.
Scenario A
Considering email exchange between R, D &D staff members which is a composite type of tacit
information to examine the rate of connectivity, direction and intensity of traffic (Howells,
2007).
Figure 1: email exchange between R, D &D staff
Table 1 indicates the number of clusters as well as the cluster constituents from the map of
research, development and design (R, D &D) staff emails.

4
Business Decision Analytics
Clusters Nodes Identified per Cluster
1 75,21, 67, 27, 15, 59, 74, 08 24
2 80, 30, 32, 70, 12, 53, 10, 54, 29, 37, 73, 58, 44, 81,
66,34, 49, 26, 68
3 19, 52, 35, 04, 36, 76, 14, 41,31, 13, 52, 45, 16, 48, 77,
38, 72, 69, 50, 47
4 83, 09, 57
5 61, 62, 07, 51, 65
6 78, 40, 39, 84
Table 1
Key members of clusters selected
Rooms Clusters Key Members Selected to
Attend Meetings
A 1 75, 74
B 2 80, 30
C 3 19, 48
D 4 83, 57
E 5 07,61
F 6 78, 40
Table 2
Type and size of each of the six clusters
Clusters are selected according to their rate of connectivity (Song & He, 2012). That is, it
is assumed that the initial node from which more connecting line originate from do indicate more
communication to other nodes. Hence the size is dictated by the extent to which the connectivity