Knowledge-based Service for African Traditional

Herbal Medicine: A Hybrid Approach

Samuel Nii Odoi Devine

, Emmanuel Awuni Kolog

, Erkki Sutinen

and Ilari Sääksjärvi

Department of Information and Communication Technology, Presbyterian University College Ghana, Okwahu, Ghana

Department of Operations and Management Information Systems, University of Ghana Business School, Accra, Ghana

Department of Future Technologies, University of Turku, Turku, Finland

Department of Biodiversity, University of Turku, Turku, Finland

Keywords: Knowledge-base, Information Retrieval, Ontology, Machine Learning, African Traditional Herbal Medicine.

Abstract: Globally, the acceptance and use of herbal and traditional medicine is on the rise. Africa, especially Ghana,

has its populace resorting to African Traditional Herbal Medicine (ATHMed) for their healthcare needs due

to its potency and accessibility. However, the practice involving its preparation and administration has come

into question. Even more daunting is the poor and inadequate documentation covering the preservation and

retrieval of knowledge on ATHMed for long-term use, resulting in invaluable healthcare knowledge being

lost. Consequently, there is the need to adopt strategies to help curtail the loss of such healthcare knowledge,

for the benefit of ATHMed stakeholders in healthcare delivery, industry and academia. This paper proposes

a hybrid-based computational knowledge framework for the preservation and retrieval of traditional herbal

medicine. By the hybrid approach, the framework proposes the use of machine learning and ontology-based

techniques. While reviewing literature to reflect the existing challenges, this paper discusses current

technologies suited to approach them. This results in a framework that embodies an ontology driven

knowledge-based system operating on a semantically annotated corpus that delivers a contextual search

pattern, geared towards a formalized, explicit preservation and retrieval mechanism for safeguarding

ATHMed knowledge.

1 INTRODUCTION

The idea of providing treatment to ailments or

diseases has, over the years, seen various methods

being applied, with each having to show its own

potency and often attributed to a specific region or

country. The case of Ghana is not different as the use

of such methods is either orthodox or traditional.

Before the introduction of orthodox medicine, many

a Ghanaian took to the use of traditional herbal

treatments, of all sorts, to help cure all forms of

diseases. Still, the use of African traditional herbal

medicine (ATHMed) is widespread in Ghana, as an

estimated 70% of the population obtain healthcare

through traditional healers (Amoah et al., 2014),

especially with rural communities. Interestingly,

preference for traditional medicine is on the rise

worldwide (Frass et al., 2012; Zhang, 2004). The

preparation of ATHMed is often in various forms

with ingredients (herbs) and methods which are

usually in the keep of the practitioners.

Predominantly and habitually, this knowledge, as

classified by Nonaka and Takeuchi (1995), is tacit,

which does not suit the view of long-term use and

preservation of such valuable knowledge. Tacit

knowledge comprises of the skills, ideas and

experiences people possess, which are hard to access

and transfer (Chugh, 2018, p. 2), in this case the

ATHMed practitioners in Ghana. There are many

cases where in an attempt to remain relevant, herbal

practitioners (traditional healers and herbalist) pass

on the knowledge to only their close allies or families.

This often leads to the likelihood of such knowledge

being either lost or failing to be developed over time

(Amoah et al., 2014). In the case where the

knowledge is not shared at all, the practitioner

eventually dies with the knowledge. Furthermore, due

to the absence of proper documentation on the choice

selection of ingredients, methods of drug preparation

and administration, some ATHMed patrons often

doubt the efficacy of these medications. There have

been reported instances of “misinformation and abuse

Devine, S., Kolog, E., Sutinen, E. and Sääksjärvi, I.

Knowledge-based Service for African Traditional Herbal Medicine: A Hybrid Approach.

DOI: 10.5220/0007946400450055

In Proceedings of the 11th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management (IC3K 2019), pages 45-55

ISBN: 978-989-758-382-7

of traditional health knowledge” (Yeboah, 2000: p.

208) which poses negative consequences to the health

of ATHMed patrons. With these concerns, this study

seeks to develop a framework for preserving and

retrieving knowledge in traditional herbal medicine.

Throughout history, attempts and varying

approaches have been developed and implemented in

order to safeguard, as well as, disseminate

knowledge. This is to ensure posterity and present

benefit from works, acts, processes, information and

all relevant data that will help improve or maintain

personal to organizational development needs.

Knowledge is regarded as key to any organization’s

present and future growth, and competitive advantage

especially in the 21st century (Xue, 2017; Davenport

and Prusak, 1998), with its basis being often formed

by, but not limited to, data and information.

To help salvage and standardize such practices in

the area of ATHMed, the United Nations Millennium

Development Goals and the World Health

Organization (WHO) has recognise the need to

promote and support the development of traditional

herbal medicine by launching the Traditional

Medicine Strategy 2014–2023 (WHA62.13)

. Key

amongst its strategic objectives is "to build the

knowledge base for active management of

Traditional, complementary and integrative medicine

through appropriate national policies". In view of this

advocacy, these researchers propose a novel model

and knowledge-based framework using

computational ontologies and machine learning to

help in the preservation and retrieval of traditional

herbal medicines. In this paper, we have suggested a

framework fit for use in Medical Institutions of

Higher Education (IHE) and herbal businesses where

the knowledge of herbal medicine, sourced from

“deep smart” knowledge bearers, is imparted to

students who are, potentially, future herbal doctors.

2 BACKGROUND

2.1 Traditional Herbal Medicine

According to the World Health Organization (WHO),

traditional medicine deals with "knowledge, skill, and

practices based on the theories, beliefs, and

experiences indigenous to different cultures, whether

explicable or not, used in the maintenance of health

as well as in the prevention, diagnosis, improvement

http://www.who.int/traditional-complementary-integrati

ve-medicine/en/

or treatment of physical and mental illness" (WHO,

2018, p. 1). Herbal medicines also encompass "herbs,

herbal materials, herbal preparations and finished

herbal products that contain as active ingredients

parts of plants, or other plant materials, or

combinations" (WHO, 2018, p. 1).

The Government of Ghana realising such a need

has setup the Centre for Scientific Research into Plant

Medicine to lead the way in the preparation and

standardization of herbal medicine in Ghana (Amoah

et al., 2014). In recent times, attempts are being made

by Universities in Ghana to train pharmacists in the

area of indigenous African herbal medicine treatment.

Two universities, Kwame Nkrumah University of

Science and Technology and University of Ghana, are

currently offering degree programmes at the

undergraduate level and training programs in an

attempt to assist in formalizing the training of

professionals in herbal medicine. However, their

focus has mainly been to help curb the challenge

related to the Ghanaian traditional herbal medicine

practitioner’s accurate measurement of ingredients

for drug preparation with issues on quality. The

universities also strive to provide strategies for long-

term preservation, appropriate forms of

administration, and administering right dosage of

such herbal drugs. For instance, the University of

Ghana

in 2016 organized a 2-day face-to-face

training programme for manufacturers of herbal

products/food supplements, focusing on the

improvement of safety and efficacy, evaluation of

raw materials, toxicological assessment, quality and

standardization in Ghana. These interventions seek to

harness the potential of herbal medicine, providing

orthodox and scientific approaches to standardizing

and safeguarding the knowledge associated with it,

and the practices in applying medication. This

attempt also assists in documenting and preserving

practices and medications associated with African

traditional herbal medicines (ATHMed) which

hitherto was in the domain of the practitioner. Failure

to undertake such interventions would in the end lead

to loss of this indigenous yet efficient medicinal

approach (Boadu and Asase, 2017).

To address the problem, a computer-aided

approach is suggested. The applications of

information technologies have been experienced in

many fields of study and are described to be a viable,

sustainable mechanism for long-term storage

(preservation) and information sharing (retrieval)

https://www.ug.edu.gh/announcements/2-day-training-

programme-manufacturers-herbal-productsfood-

supplements

KMIS 2019 - 11th International Conference on Knowledge Management and Information Systems

resident in the domain of knowledge management. To

this end, a knowledge-based framework for

preserving and retrieving ATHMed in the service

pharmaceutical sector and assistive of formal learning

in Medical Institutions of Higher Education is

proposed. The framework provides a roadmap on

describing some key elements such as choice of

medicinal plants, relationship between them and the

associated diseases they seek to cure. This would

provide the base for preserving the knowledge of

traditional herbal medicine practitioners (formal and

informal), thus, the engineering of an African

Traditional Herbal Medicine ontology. This is geared

towards assisting in the efficient classification and

annotation of the knowledge of treatment in

traditional herbal medicine, in order to build the

relevant contextual relationship between ATHMed

remedies and their associated cures. The fulcrum of

designing such a retrieval mechanism is to enable

efficient access to the knowledge base (KB). A

software artefact developed based on the framework

is needed to test the adequacy and viability of the KB.

By this approach, the knowledge preservation of

ATHMed will aid in providing a formalized, explicit

repository for future pharmaceutical needs and formal

training.

2.2 Related Works in Knowledge-base

Archival

Since the 1990s, with the advent of computers,

knowledge management (KM) has been implemented

through software solutions. This is especially in the

employee training, based on well documented norms,

practices and personal know-how (tacit knowledge)

of other internal employees, so that sharing and

safeguarding knowledge can be efficiently done

(Davenport and Völpel, 2001; El Morr, 2010).

Extant literature has revealed that, preservation of

data and knowledge has been the focus of many

organizations since the integration of procedures to

manage and improve information assets has become

critical for competitive growth in the 21st century

(Xue, 2017). However, the preservation of

knowledge, especially tacit, is hard to transfer and

difficult to preserve (Mazour, 2006). By implication,

there is the need for a more robust, explicit and

systematic approach to capturing, classifying and

sharing such knowledge. The involvement of

computerization and various electronic-based

strategies also provide additional power to monitor

and manage electronic information (Davidavičienė

and Raudeliūnienė, 2010).

Ashkenas (2013) advocates that to sustain,

adequately formalize, explicitly define and ensure

continuity of institutional knowledge, information

technology, thus computational approaches, should

be adopted. In agreement, Panahi, Watson and

Partridge (2012) espouse that using computational

approaches are more effective and efficient than

verbal/oral or face-to-face approaches. As knowledge

sharing, which hinges on efficient knowledge

preservation (KP), particularly tacit knowledge, is

paramount and unavoidable (Sarkiunaite and

Kriksciuniene, 2005), this paper proposes a hybrid

approach to KB implementation for ATHMed, based

on prior approaches adopted in other related fields.

Efficient and accurate retrieval of knowledge is

heavily dependent on how it is stored, herein,

preservation. The relevance and critical role of

knowledge preservation cannot be downplayed as it

drives the very essence of continuous learning and

improvement for posterity (Faust, 2007).

Knowledge preservation can be viewed as

“a process for maintaining knowledge important

to an organization’s mission that stores

knowledge/information over time and provides the

possibility of recall for the future” (Mazour, 2006: p.

2). Probst et al., (2006) also argue that the processes

of KP are covered in three stages: select, store,

actualize. In an extensive review undertaken by

Antonova et al., (2006) on technology solutions for

KM, the researchers explored numerous approaches

employed within the domain. Antonova et al., (2006)

detail notable technological solutions in managing

knowledge by classifying them based on the key KM

processes encompassing: ‘generation’, ‘storing,

codification and representation’, ‘transformation

and use’ and ‘transfer, sharing, retrieval, access and

search’. This paper discusses KP and retrieval,

focusing on storage, codification and representation

of knowledge (Antonova et al., 2006), considering

KP as a purposeful, articulate, and explicit activity

that involves the safeguarding of knowledge,

especially tacit knowledge, for efficient storage,

retrieval, use and dissemination.

Over the years, there has been a study migration

from merely employing information system strategies

that mainly focused on transaction processing,

process controls and assisting in decision making, to

the efficient capturing and dissemination of

interrelated knowledge. This shift has become a vital

component to many an organization’s operations.

This focus seeks to guarantee the future success and

sustenance of organizations, using relevant,

progressive and adaptive computational approaches.

Mazour (2006) emphasizes that documentation,

Knowledge-based Service for African Traditional Herbal Medicine: A Hybrid Approach

hitherto preservation, is a “good means” to “articulate

knowledge” and though knowledge (tacit)

preservation and retrieval involves extensive effort,

its benefit is undeniably vast.

Based on the classification proposed by Probst et

al. (2006) for tacit knowledge preservation processes,

Davidaviciene and Raudeliuniene (2010), in an

extended work, identified ICT tools that implement

the KP process. The implementation is carried out

through capturing systems (expert systems, chat

groups, wikis, blogs, podcasting, best practices and

lessons learned databases, computer based

communication, computer based simulation) and

sharing systems (team collaboration tools, wikis,

blogs, podcasting, web-based access to data,

databases, best practice databases, lessons learned

systems and expertise locator systems)

(Davidaviciene and Raudeliuniene, 2010).

Antonova et al., (2006) catalogued technological

solutions for storage of knowledge assets into

databases, knowledge bases, data warehouses and

knowledge warehouses, data marts and data

repositories. With regards to knowledge codification

and representation, case-based reasoning systems,

rule-based approaches, frame and semantic nets as

well as formal logical, production and procedural

model approaches are engaged (Antonova et al.,

2006). Antonova et al. (2006) further postulate that

for retrieval to be easily facilitated, knowledge

organization technologies which include directories,

taxonomies and repository indexes must be adopted.

This suggests the adoption of ontology-based design

specifications. In addition to these technologies for

knowledge storage and representation, topic and skill

maps, and controlled vocabularies, structured as data

dictionaries can be applied (Antonova et al., 2006).

Presently, research leans towards contextual

semantic preservation of knowledge assets using

semantic enabling technologies such as XML. The

use of XML has aided the provision of interrelated

pattern search in a platform-independent

environment, as evident in the web environment.

Other studies have indicated the integration of natural

language processing (NLP) techniques to provide

linguistic semantic analysis and pattern related search

and retrieval of knowledge (Tomai and Spanaki,

2005). Additionally, machine learning techniques for

intuitive extraction of knowledge is being adopted.

As earlier espoused, KB development related

literature indicates focus towards intuitive, robust,

platform-independent, metadata dependent,

semantically driven, contextual domain-based

knowledge modelling, reasoning, preservation and

extraction systems. A number of these computational

approaches have been explored and adopted over the

years.

In their study, Li, et al., (2003) investigated the

feasibility of adopting an ontology-based approach to

supporting KM activities in the metal industry in

Taiwan. The researchers argue that, though this

approach is an “adequate methodology” for domain

specific KM, it is rather involving and time-

consuming. To this end, an ontology-based KMS was

modelled, using KAON for building the ontology and

Java 2 Enterprise Edition (J2EE) for the system in a

web-based environment.

Similarly, Tomai and Spanaki (2005) adopted the

use of ontologies to propose a framework for

modelling robust geographic concepts. However,

they utilized NLP approaches for linguistic semantic

analysis of the knowledge in the KB. To enable data

be appropriately tagged, in context, the metadata

handling and description language, Web Ontology

Language (OWL) full was used. This enabled the

application of W3C specification tools, thus

supporting Extensible Markup Language

(XML)/XML Schema and Resource Definition

Framework (RDF)/RDF Schema. A web-based

environment for user interaction with the KB was also

adopted, built on the .ASP format, and tested in an

intranet.

Lin, et al., (2013) also addressed the challenge of

tackling environmental protection concerns in the

product development process, through an eco-design

approach, using a Life Cycle Assessment (LCA)

strategy. The researchers presented an ontology-

based process-oriented framework as the

underpinning strategy to address the problem, with

Protégé as the tool for building the ontology. A

reasoning engine was formed to establish an

ontology-based knowledge decision support system

by transforming the knowledge acquired into a

linguistic information, fit for OWL and Semantic

Web Rule Language (SWRL), using the Java Expert

System Shell (Jess).

In the aerospace industry domain, Sanya and

Shehab (2014) propose a novel approach towards the

development of a KB engineering framework for

implementing platform-independent knowledge-

enabled product design systems. With the aim of

achieving long-term preservation of knowledge on

engineering, their framework targeted a cost-

effective, reasonable, model-driven architecture, also

using an ontology-based approach that renders a

semantic and portable knowledge structuring system.

OWL and SWRL were used to model the KB system,

Java for the CAD interface and Jess was used for the

inference engine.

KMIS 2019 - 11th International Conference on Knowledge Management and Information Systems

Song et al., (2016) explored a 3-tier system

architectural framework for KB systems’

management of manufacturing process knowledge.

The researchers proposed an effective reusable

management approach to implementing their KB

system, via a systematic methodology for

constructing the KB, indicating the key role of

ontology development through an iterative process.

According to Shang et al., (2017), by employing a

vulnerability-centric ontology-based KB framework

strategy, cyber security knowledge existing in some

independent KBs and on the internet, in text form,

was efficiently integrated. This enabled the extraction

of cyber security knowledge using both rule-based

and machine learning information extraction

techniques.

From the literature reviewed, in addressing KM

needs, a systematic approach is to be adopted. The

approach should be dynamic, contextual, ontology-

based, AI-oriented, platform independent and

semantic in nature. This forms the basis of the design

and development approach adopted for this research.

3 METHODOLGY

This research is an ongoing study that adopts a design

science research process. In this section, we provide

the overview of the DSR process and the framework

for the implementation of the system.

3.1 Design Science Research

As part of this project, the development of the

knowledge-based system follows a design science

research (DSR) process. Peffers et al. (2006) DSR is

adopted as shown in Figure 1. DSR is one of the two

major paradigms in Information System research

(Hevner et al., 2004; Mramba et al., 2016) and is a

solution-oriented methodology that focuses on the

design and investigation of IT artefacts (Peffer et al.,

2006; Hevner et al., 2004). DSR seeks the systematic

and purposeful development of a solution (artefact)

based on a rigorous scientific process. It provides a

strategic mental model for empirical and theory

research building for contextual understanding,

evaluation and reproducibility of a study. The artefact

may take the form of a construct, model, method, or

an instantiation such as hardware or software (Hevner

et al., 2004).

For an Information Systems’ project such as this

study, Peffer et al., (2006) propose six common

design process elements that cover DSR methodology

in a nominal sequence: Problem Identification and

Motivation, Objectives of a Solution, Design and

Development, Demonstration, Evaluation and

Communication. This study takes a problem-centered

focus as its entry point. Consequently, the DSR

methodology is well suited for this research as our

aim is to develop an artefact that provides a formal

computational solution for safeguarding ATHMed

knowledge (Peffers et al., 2006). Subject to this

approach, stakeholders for this project will be

involved throughout the developmental stages as

shown in Figure 1. The idea is to co-create the

intended system with the stakeholders’ involvement.

This paper forms part of the initial phase of the

DSR framework. Based on the DSR’s initial phase,

this study identifies the gap in the knowledge

preservation and retrieval (KPR) practices within the

ATHMed domain. From literature, it is evident that

the KPR practices are inadequate with its

repercussion having been discussed in earlier sections.

The motivation therefore is to propose a framework

to assist in addressing the phenomenon. This will

require clearly defined requirements to help obtain

relevant information and guidelines needed for

identifying key concepts towards the KBS

development.

Of the requirements specified through active

stakeholders’ consultation, the artefact will be

designed and developed towards meeting the set

objectives. Four components of the artefact will be

materialized to provide a contextually fit and rigorous

solution. The artefacts are interdependent covering an

ATHMed Framework to guide the development

process, a formalized ontology, a corpus and a KBS.

The prototyping strategy shall be used to enable

stepwise involvement of stakeholders. As a proof of

concept, the KBS will undergo rigorous testing with

training data to demonstrate its ability and adequacy

to meet the objective for which the system was

designed for.

Subsequently, the output, efficiency and overall

performance of the artefact will be measured through

another stage of rigorous testing by stakeholders and

patrons of ATHMed. This will help in assessing the

veracity and efficacy of the KB and how usable the

KBS is. The feedback obtained will affirm system

adequacy and viability, as well as usability and user

satisfaction.

Knowledge-based Service for African Traditional Herbal Medicine: A Hybrid Approach

Figure 1: Design process with design science framework (Adopted from Peffers et al., 2006).

3.2 Knowledge-based Framework for

Traditional Herbal Medicine

To contribute to the building of a formally defined

body of knowledge in the domain of ATHMed

practice and training in Medical IHE, a conceptual

framework (model) is proposed. As such, defining a

semantic-oriented, lexically functional, syntactic

structure in a machine-readable and machine-

understandable language is the focus of this proposed

framework. The framework seeks to adopt and

incorporate two major Artificial Intelligence (AI)

techniques – machine learning (ML) and ontology– to

provide a progressively adaptive environment to

implement an intuitive KBS. Additionally, the

retrieval mechanism for users’ interaction is

facilitated using web and mobile technologies. The

proposed knowledge-based framework has three

components as shown in Figure 2: data layer, logic

layer and application layer. The system will be

(MVC) software architecture. The Data Layer

implements the Model component, the Logic Layer

actualizes the Control component and the View

component is considered in the Application Layer.

3.2.1 The Data Layer

The data layer is comprised of the data (knowledge)

acquisition sub-component, which captures the

knowledge, and knowledge base, which stores the

knowledge obtained from the domain experts.

Data Processing

As earlier indicated, with the DSR, stakeholders in

the medical fraternity are involve in the actual

implementation. Therefore, the initial data (ATHMed

corpus) obtained from interacting with the

stakeholders (ATHMed practitioners and academia)

is to be pre-processed for easy annotation. This

processing is based on the elements relevant for

ingredient selection and processing, mode of

preparation and associated treatment (administration)

method. Through expert (experts from academia,

research institutions and registered herbal clinics)

annotation, a predefine annotation scheme for the

ATHMed data shall be arrived at. This will enable the

attainment of a good agreement score on what

element goes where, to obtain a concisely annotated

data for training a ML classifier. This will lead to a

gold standard being obtained, from the initial test

data, via a supervised machine learning technique.

During storage, which implements the

preservation phase, the data after pre-processing, is

run through a classifier for categorization into

specific non-static association and accurate

identification of the knowledge using a suitable ML

classifier. Throughout classification process, each

data value (or key words) will be tagged to enable

efficient association of term and concepts, via part-

of-speech tagging through lemmatization.

Literature in medical research involving the use

of ML suggests a diversity of techniques being

applied. Jiang et al., (2017) conducted a survey of AI

KMIS 2019 - 11th International Conference on Knowledge Management and Information Systems

application in healthcare. The researchers observed

that Support Vector Machine (SVM) and Artificial

Neural Networks (ANNs) were often used. In a

related study on biomedical disease detection,

Niharika and Kaushik (2018) concluded that different

ML classifiers should be considered to build a unified

hybrid framework. By this, to obtain optimal results,

they recommend the use of SVM, extreme learning

with variations of various swarm techniques.

Similarly, Bayesian Networks (BNs), ANNs and

SVM were identified as methods most frequently

applied in the THMed domain (Arji et al., 2018).

Consequently, to determine which ML algorithm

is most suited for the ATHMed classifier, an optimal

performance test shall be undertaken. This test shall

involve the use of notable machine learning software,

example WEKA. Based on the results, ML that

demonstrate superiority in classifying ATHMed data

will be chosen for the implementation.

Additionally, a robust algorithm to improve

classification of the ATHMed data shall be

considered. This is to directly influence the

performance of ontology model, through a novel

indexing approach, to facilitate quick retrieval.

The Ontology Model

The proposed knowledge base (KB) design is based

on an ontology, for extensibility and reusability. As

such, at the ontology construction phase, the proposed

ontology model is to be defined through the process

of organizing the ATHMed knowledge in categories,

type classification into classes, hierarchical

structuring of associated instances and properties,

with appropriate axioms, rules, restrictions and

interrelations. The ontology model is to be evaluated

to verify functional specifications thereby ensuring

consistency, reliability, accuracy and extensibility. A

further validation, involving verification and

evaluation tests for reliability (Preece, 1994) is

required to ascertain the performance capacity of the

KBS. This is to measure the quality of the ontology

by ensuring adequate coverage of the knowledge

within the ATHMed domain. The testing strategy for

the KB is tailored to conform to software

development test approaches.

The ontology-based approach is predominantly

novel yet extensive for knowledge engineering. It

facilitates appropriate representation, maintenance

and dissemination of knowledge, and thus relevant in

applying this approach in the ATHMed domain. The

proposed knowledge capturing/acquisition

component is targeted at being modelled based on a

standard medicine preparation template design focus,

under the guide of a domain expert (in academia) and

verified by experts (both academia and practitioners).

This shall require the use of knowledge and

metadata representation tools. An ontology authoring

tool such as Protégé shall be used for building the

ontology. XML and XML Schema is suitable for

defining the metadata description, structure and

storage, with RDF for describing related knowledge

resources prevalent on the web for interrelation

definition. OWL Descriptive Logic (DL) shall be

used for defining the ontology structure and necessary

semantics, and SPARQL to querying the ontology-

based knowledge-base. A classification of key

elements and concepts, with their interrelationships

shall be explored, defined and annotated. Appropriate

axioms shall be used to implement the guidelines for

building the ontology.

Through this approach, the proposed knowledge-

based system is certain in providing knowledge users

and ATHMed stakeholders with a computer-aided

solution that supports semantic capabilities. This will

facilitate relevant, contextual preservation and

retrieval of results from the knowledge base, through

a simplified search mechanism using a web portal or

mobile app.

3.2.2 The Logical Layer

This layer, as the middle layer, interacts with the data

layer where the knowledge is preserved and the

application layer, where requests from users are

submitted via queries. Thus, the reasoning module is

executed via the logic layer, tagged as the Information

Extraction Component (IE). The IE is implemented

via two sub-components: Machine Learning

Extractor and Semantic Reasoner.

Machine Learning Extractor

The information extraction component shall be used

as the means to extract or retrieve information from

the knowledge base. The IE constructed using two

sub-approaches involving a machine learning based

(ML-based) extraction algorithm and an inference

engine. The ML extractor will be designed to

complement the ontology, that possess features of the

ATHMed knowledge-based ontology, taking critical

considerations to class types, their relationships,

instances, connotations and semantics. The ML

extractor interacts with the inference engine built to

deduce new knowledge, thus from the data and

information present in the knowledge base (KB) to

ensure accurate prediction and determination of

Knowledge-based Service for African Traditional Herbal Medicine: A Hybrid Approach

appropriate treatment, association to ailment,

preparation methods and other relevant inferences.

Semantic Reasoner

The proposed knowledge-based system’s semantic

reasoner acts as the inference engine. The inference

engine structure is expected to work on the Java-

based framework, Apache Jena. ML techniques

mainly to improve efficiency of predictability,

deduction accuracy and an extensible deeper learning

approach shall be considered. This shall support

making of inferences whiles deducing new

knowledge in a contextual form. This requires that the

ML algorithm (classifier) chosen must be trained,

firstly on some rule-based approach before a well-

structured corpus for ATHMed is defined and

extended with the algorithm.

3.2.3 The Application Layer

As earlier mentioned, the application layer is the view

component, thus presentation layer for accessing the

KBS. In the wake of enhanced and fluid ways of

communication, there is the need to make access to

knowledge easy and quick, supported in an

interconnected, platform independent environment.

This requires employing technologies that are

ubiquitous and support use, anywhere and at anytime

with ease. Web-based and mobile technologies have

shown potential to be adaptive to such an

environment. The proposed KBS is targeted at

providing users with a graphical front-end: web-

based interface designed with J2EE technology and a

mobile interface using Android technology. Users

will pass queries through simple to expressive

statements which are treated as natural language

expressions. NLP techniques and tools are efficient to

support such manipulation. This requires that all

queries be analysed lexically, and semantically

compared against the KB through the information

extraction (IE) component. This is to facilitate the

extraction of relevant knowledge in the form required

for adequate interpretation and understanding to the

user. To this end, representation of the extracted

knowledge must be textual and visual where required.

Same interface shall be used to provide new

knowledge to the KB.

For optimal performance, thus reliability and

accuracy of the proposed KB, systematic engagement

of domain experts in designing and validating every

stage of the KBS is required. Below is a conceptual

design of the proposed ATHMed KB framework.

Figure 2: The Proposed knowledge-based framework for Achieving African traditional herbal medicine.

KMIS 2019 - 11th International Conference on Knowledge Management and Information Systems

4 DISCUSSION

Developing countries have about 80% of their

populace depending on THMed, with reports of rising

global demands, for their health needs (WHO, 2018;

Frass et al., 2012; Zhang, 2004). This is due to

THMed’s ease of use, natural form, often minimal

side effects, availability, affordability and ease of

preparation. Poorna, Mymoon and Hariharan (2014)

reported cases of countries pursuing knowledge

preservation and retrieval practices on traditional

knowledge, thus THMed. Their findings, indicate

strategic, meticulous efforts to safeguard THMed

knowledge, have yielded distinctive yet

collaboratively immense benefits. This result was

attained through an all-inclusive effort spanning the

health, academic and industry divide. Interestingly,

all 6 countries that were observed to be making strives

by implementing THMed documentation initiatives

were from developing nations. This reiterates the

importance of THMed to developing nations, and

African countries in particular.

In essence, the impact of pursuing such THMed

knowledge preservation and retrieval leads to key

benefits to countries especially in Africa. There are

obvious implications of healthcare benefits. It provides

economic benefits whiles promoting preservation of

indigenous cultural heritage (Boadu and Asase, 2017;

van Andel et al., 2015). It provides avenues for

securing patents to THMed knowledge (Zhang, 2004)

and protecting such knowledge against misuse

(Poorna, et al., 2014). These benefits were witnessed

amongst the 6 countries that were observed by Poorna,

Mymoon and Hariharan (2014). This was possible

through proper codification and documentation

(Poorna, Mymoon and Hariharan, 2014), a requisite

strongly advocated for, to facilitate preservation

(Faust, 2007; Mazour, 2006). This means that for

THMed knowledge to be preserved, a consistent and

purposeful effort is required. In addition, it is necessary

that all such knowledge captured are interlinked or

semantically correlated, to enable efficient search and

retrieval, in order to promote new discoveries in related

medicines and treatments.

Despite extant literature indicating growing

interest in herbal medicine usage, especially in

developing countries, most of these practices are

either not well preserved nor documented. The

evidence and dangers of THMed knowledge not

being properly and adequately transferred from one

bearer to another, to continue the medicinal practice

is well documented in literature. Instances of this

nature were observed by Adekannbi, et al., (2014).

The researchers indicate that, though some

practitioners of THMed willingly share their

knowledge with their assistants, they choose to

purposefully, deliberately and prudently transfer such

knowledge, comfortably with their own relations.

Notably, albeit some African countries are attempting

to safeguard THMed knowledge, it is predominantly

oral and being leisurely done using some form of

recording (Maluleka and Ngulube, 2018; Adekannbi,

et al., 2014; Yeboah, 2000). Consequently, a

knowledge bearer may teach in part resulting in half-

baked practitioners administering poor medical

practices which are likely to cause more harm than

good, as evidenced in some reports (Yeboah, 2000).

To help salvage the situation, deliberate calls and

attempts have been made, by academia and research

institutions worldwide, to help formalize and train

THMed professionals. As custodians, these

professionals will not only guide the processes and

procedures to the preparation, preservation and

administration of THMed, but also ensure the

knowledge is not lost (WHO, 2018; Boadu and Asase,

2017; Amoah et al., 2014; Poorna, et al., 2014).

Additionally, this effort is to assist in safeguarding

such vital knowledge which can be extracted and

refined for further commercialization in the service

industry. Key to benefiting from this are researchers

and academia, in general, via extending studies into

the viability of such medicines and the development

of appropriate curricula in the training of qualified

herbal medicinal practitioners and healthcare

professionals who will ensure continuity and

formalization of THMed practices in Higher Learning

Institutions. The resulting output from academia will

feed the pharmaceutical industry with the requisite

knowledge on THMed. Causally, it will provide

opportunities for job creation while sustaining

economies of countries that depend predominantly on

THMed for providing healthcare to their indigenes.

Ultimately, the effort will assist in improving socio-

economic livelihoods and sustaining biological

resources at the same time. The overall effect shall be

to the advantage of the general populace with regards

to the provision of relevant information on what

choice of THMed medication to resort to in servicing

their medical needs. The study conducted by Poorna,

et al., (2014) supports these claims. In their study, it

was revealed that countries that pursued preservation

of THMed knowledge benefited not only to safeguard

such knowledge, but also inherently promoted saving

and securing patent to these national assets.

The provision of improved, timely, accessible and

affordable healthcare through the offering of

alternative healthcare practices based on a formalized

THMed knowledge-based Framework is the

Knowledge-based Service for African Traditional Herbal Medicine: A Hybrid Approach

underlining motivation proposed in this paper. The

expected impact will not only be for developing

countries like Ghana, but all regions that value and

access healthcare through THMed.

The proposed framework is geared towards

managing African traditional herbal medicine

(ATHMed) knowledge, focusing on its relevance in

the Ghanaian pharmaceutical service sector and the

medical institutions of higher education context,

seeking to deepen research at the national and global

level, as advocated by Xu et al. (2008). The

framework falls in the domain of KM processes

covering technologies, related tools, appropriate

methods and general KM involving frameworks.

Furthermore, the proposed framework seeks to add to

the KM body of knowledge, by contributing to the

delivery and improvement of quality healthcare (El

Morr and Subercaze, 2010). This, we hope, will

extend knowledge in ATHMed and contribute in

answering the call for preserving such indigenous

medicinal practices (WHO, 2018).

5 CONCLUSION AND FUTURE

WORK

The premise and prospects of this work is based on

the articulation of formalized processes that are

expected to yield a structured computer science

approach to assist in preserving and retrieving

relevant data and information on African Traditional

Herbal Medicine, progressively, efficiently and

sustainably, using novel approaches.

A hybrid approach of adopting machine learning

techniques and an ontology framework is presented.

This is geared towards addressing issues related to

developing an efficient knowledge-based system

(KBS), that provides a formal, explicit preservation

and retrieval mechanism of ATHMed leading to a

well-defined body of ATHMed knowledge.

This paper is considerably a working paper,

requiring further empirical studies to concretize and

affirm the veracity and viability of the framework. In

this regard, further research will focus on the design

of an ontology, based on ATHMed. Trailing this will

be the exploration of strategies to facilitate quick and

adept cataloging of data through an innovative search

pattern that employs a novel indexing approach to

storage and data retrieval, applicable to classifying

semantic data, thus an annotated corpus. Furthermore,

research into the adoption of machine learning

techniques tailored to suit annotated ATHMed

knowledge to provide a semantically-aware,

contextual knowledge base shall be pursued. The

research then culminates with an implementation of

the framework proposed and validation of the

ATHMed KBS.

REFERENCES

Adekannbi, J., Olatokun, W. M., Ajiferuke, I., 2014.

Preserving traditional medical knowledge through

modes of transmission: A post-positivist enquiry. South

African Journal of Inform Management, 16(1), pp.1-9.

Alavi, M., Leidner, D. E., 2001. Knowledge management

and knowledge management systems: Conceptual

foundations and research issues. MIS quarterly, Vol. 25

(1), pp. 107-136.

Amoah, S. K., Sandjo, L. P., Bazzo, M. L., Leite, S. N.,

Biavatti, M. W., 2014. Herbalists, traditional healers

and pharmacists: a view of the tuberculosis in Ghana.

Brazilian Journal of Pharmacognosy, 89-95.

Antonova, A., Gourova, E. and Nikolov, R., 2006, July.

Review of technology solutions for knowledge

management. In 2006 2nd IET International

Conference on Intelligent Environments-IE 06 (Vol. 2,

pp. 39-44). IET.

Arji, G., Safdari, R., Rezaeizadeh, H., Abbassian, A.,

Mokhtaran, M., Ayati, M. H., 2019. A systematic

literature review and classification of knowledge

discovery in traditional medicine. Computer methods

and programs in biomedicine, 168, pp.39-57.

Ashkenas, R., 2013. How to preserve institutional

knowledge. Harvard Business Review.

Boadu, A. A., Asase, A., 2017. Documentation of herbal

medicines used for the treatment and management of

human diseases by some communities in southern

Ghana. Evidence-Based Complementary and

Alternative Medicine, 2017.

Bolisani, E., Bratianu, C., 2018. The Elusive Definition of

Knowledge. In Emergent Knowledge Strategies (pp. 1-

22). Springer, Cham. DOI: https://doi.org/10.1007/978-

3-319-60657-6_1

Carrillo, F. J., 2006. From transitional to radical

knowledge-based development. Journal of Knowledge

Management, 10(5), 3-5. DOI: https://doi.org/

10.1108/13673270610691125

Chugh, R., 2018. Tacit Knowledge Transfer in Australian

Universities: Exploring the Barriers and Enablers. In

MATEC Web of Conferences (Vol. 210, p. 04054).

EDP Sciences. DOI: https://doi.org/10.1051/matec

conf/201821004054

Davenport, T. H., Prusak, L., 1998. Working knowledge:

How organizations manage what they know. Harvard

Business Press.

Davenport, T. H., Völpel, S. C., 2001. The rise of

knowledge towards attention management. Journal of

knowledge management, 5(3), 212-222.

Davidavičienė, V., Raudeliūnienė, J., 2010. ICT in tacit

knowledge preservation. In Proceedings of the 6th

KMIS 2019 - 11th International Conference on Knowledge Management and Information Systems

International Scientific Conference “Business and

Management 2010”, Selected Papers (pp. 436–442).

El Morr, C., Subercaze, J., 2010. Knowledge management

in healthcare. In Handbook of research on develop-

ments in e-health and telemedicine: Technological and

social perspectives (pp. 490-510). IGI Global.

Faust, B., 2007. Implementation of tacit knowledge

preservation and transfer methods. In International

Conference on Knowledge Management in Nuclear

Facilities (pp. 18-21).

Frass, M., Strassl, R. P., Friehs, H., Müllner, M., Kundi, M.,

Kaye, A. D., 2012. Use and acceptance of

complementary and alternative medicine among the

general population and medical personnel: a systematic

review. The Ochsner Journal, 12(1), 45-56.

Girard, J., Girard, J., 2015. Defining knowledge

management: Toward an applied compendium. Online

Journal of Applied Knowledge Manag., 3(1), 1-20.

Hevner, A. R., March, S. T., Park, J., Ram, S., 2004. Design

science in information systems research. MIS quarterly,

28(1), 75-105.

Jiang, F., Jiang, Y., Zhi, H., Dong, Y., Li, H., Ma, S., Wang,

Y., Dong, Q., Shen, H., Wang, Y., 2017. Artificial

intelligence in healthcare: past, present and future.

Stroke and vascular neurology, 2(4), pp.230-243.

Li, S. T., Hsieh, H. C., Sun, I. W., 2003. An Ontology-based

Knowledge Management System for the Metal

Industry. In WWW (Alternate Paper Tracks).

Lin, J. S., Hsu, W. L., Chang, J. H., 2013. An ontology-

based product development framework considering

eco-design. In Proceedings of the International Multi

Conference of Engineers and Computer Scientists.

Maluleka, J. R., Ngulube, P., 2018. The preservation of

knowledge of traditional healing in the Limpopo

province of South Africa. Inform. Dev, 34(5), 515-525.

Mazour, T., 2006. Knowledge preservation and transfer:

issues and terminology. Proceedings of School of

Nuclear Knowledge Management, pp.18-22.

Mramba, N., Apiola, M., Kolog, E.A. and Sutinen, E., 2016.

Technology for street traders in Tanzania: a design

science research approach. African Journal of Science,

Techn, Innovation and Development, 8(1), pp.121-133.

Nonaka, I., Takeuchi, H., 1995. The knowledge-creating

company: How Japanese companies create the

dynamics of innovation. Oxford university press.

Niharika, Kaushik, B. N., 2018. Machine Learning in

Biomedical Mining for Disease Detection. Journal of

Artificial Intelligence, 11: 39-47.

Panahi, S., Watson, J., Partridge, H., 2012. Social media

and tacit knowledge sharing: Developing a conceptual

model. World academy of science, engineering and

technology, (64), pp.1095-1102.

Peffers, K., Tuunanen, T., Rothenberger, M. A. Chatterjee,

S., 2006. A Design Science Research Methodology for

Information Systems Research. Journal of

Management Information Systems, 24(3), Winter 2007-

8, 45-78.

Poorna, R., Mymoon, M., Hariharan, A., 2014. Preserva-

tion and protection of traditional knowledge – diverse

documentation initiatives across the globe. Current

Science, 107(8):1240–1246.

Preece, A., 1994. Validation of knowledge-based systems:

The state-of-the-art in north america. Journal of

communication and cognition-Artificial intelligence,

11, 381-413.

Probst, G., Raub, S., Romhardt, K., 2006. Ziniu vadyba:

sekmes komponentai [Managing Knowledge: Building

Blocks for Success]. Vilnius: Knygiai. 352 p. ISBN

9955-443-26-X

Sanya, I. O., Shehab, E. M., 2014. An ontology framework

for developing platform-independent knowledge-based

engineering systems in the aerospace industry.

International Journal of Production Research, 52(20),

6192-6215.

Sarkiunaite, I. D., Kriksciuniene. 2005. Impacts of

information technologies to tacit knowledge sharing:

Empirical approach. Informacijos mokslai, pp. 69-79.

Shang, H., Jiang, R., Li, A., Wang, W., 2017. A Framework

to Construct Knowledge Base for Cyber Security.

In Data Science in Cyberspace (DSC), 2017 IEEE

Second International Conference on (pp. 242-248).

IEEE.

Tomai, E., Spanaki, M., 2005. From ontology design to

ontology implementation: A web tool for building

geographic ontologies.

Song, H., Wang, H., Liu, T., Zhang, Q., Gao, B., 2016. The

Design and Development of Manufacturing Process

Knowledge Base System Based on Ontology. In

International Conference on Cooperative Design,

Visualization and Eng. (pp. 9-16). Springer, Cham.

Van Andel, T. R., Croft, S., van Loon, E. E., Quiroz, D.,

Towns, A. M., Raes, N., 2015. Prioritizing West

African medicinal plants for conservation and

sustainable extraction studies based on market surveys

and species distribution models. Biological

Conservation, 181, 173–181.

World Health Organization, 2018. Traditional,

complementary and integrative medicine. [online]

Available at: http://www.who.int/traditional-comple

mentary-integrative-medicine/about/en/ [Accessed 18

Feb, 2018]

Zhang, X., 2004. Traditional Medicine: Its Importance and

Protection. Protecting and Promoting Traditional

Knowledge: Systems, National Experiences and

International Dimensions. Part, 1, 3-6. Available at:

https://unctad.org/en/docs/ditcted10_en.pdf#page=23.

Xu, J., Sankaran, G., Sankaran, S., Clarke, D., 2008.

Knowledge management in twenty-first century:

Literature review and future research directions. The

Intern. Technology Management Review, 1(2), 16-24.

Xue, C. T. S., 2017. A Literature Review on Knowledge

Management in Organizations. Research in Business

and Management, 4(1), pp.30-41.

Yeboah, T., 2000. Improving the provision of traditional

health knowledge for rural communities in Ghana.

Health Information & Libraries Journal, 17(4), 203-

208.

Knowledge-based Service for African Traditional Herbal Medicine: A Hybrid Approach