AN APPLICATION OF ANT COLONY OPTIMIZATION IN A KNOWLEDGE MINING ENVIRONMENT

International Journal of Computer Engineering and Applications, Volume VII, Issue II, Part II,

August 14

222 P. S. Ratnaparkhi and P. K. Butey

AN APPLICATION OF ANT COLONY OPTIMIZATION IN A

KNOWLEDGE MINING ENVIRONMENT

Ms. Prajakta S. Ratnaparkhi 1, Dr. Pradeep K. Butey

1 Research Scholar,

Department of Electronics & Computer Science, RTM Nagpur University, Nagpur

2 Research Supervisor, Department of Computer Science, Kamla Nehru Mahavidyalaya, Nagpur

ABSTRACT:

In today’s globalized economy particularly under the pressure of economic challenges and the uncertainty that organization face while making decisions has a significant impact on economic growth of any organization. Key factor in the success of the organization depends on its ability to process the data and effective handling of uncertainty. Many Fortune 500 companies are recognizing enterprise data as a strategic business asset. Leading companies are striving hard to optimize their business processes and create intelligent product to satisfy their demands and survival of the organization in such a competitive environment. Large companies struggle to access, manage and leverage the information that they create in their organization as well as their experts working for so long period they are full of information and wisdom organization is striving hard to manage this tacit data in most effective manner. This is what leads to growing demand of knowledge management. Every organization wants to take better decisions as fast as they can. Hence the ability to take advantage of growing amount of knowledge that organization has become an extremely critical factor for the success of any organization. That is again the reason for global demand for effective knowledge management system. Knowledge warehouse is the repository where all the tacit knowledge is transferred in an explicit form and is updated regularly by the knowledge worker. In this paper we are discussing about use of ant colony optimization optimum selection in the most uncertain condition we are conceiving that the required data is already stored in Knowledge Warehouse.

Keywords: Ant colony optimization, business intelligence, data warehouse, knowledge

management system, Knowledge worker, optimization techniques.

[1] INTRODUCTION

Knowledge, as now everybody say, is power. It is also one of the business’s most vital

forms of capital – the bedrock on which innovation is founded, and the fuel that drives growth

and expansion. But many companies pay insufficient attention to how knowledge is captured,

nurtured, shared and perpetuated. This can lead to the degradation of a business’s information

resources, and a failure to maximize employee potential. In order to ensure this does not happen

to your business, a robust and dynamic knowledge management strategy is vital.

It is generally agreed by IT practitioners that there exists a continuum of data,

information, and knowledge (and even wisdom) within any enterprise. [1]. Now a day’s people

An Application Of Ant Colony Optimization In A Knowledge Mining Environment


are trying to codify this wisdom and reap the benefit in long term so as to sustain in this

challenging world. Due to this study of knowledge management is becoming very popular and

researchers are trying to develop many tools to support quick and effective decision making.

According to these researchers, data consist of facts, images, or sounds. When data are

combined with interpretation and meaning, information emerges. This overload of data is

making knowledge management increasingly more important. Three key reasons why actively

managing knowledge is important to a company’s success are:

1.) Facilitates decision-making capabilities, 2.) Builds learning organizations by

making learning routine, and, 3.) Stimulates cultural change and innovation. All this

information is formatted, filtered, and summarized data that, when combined with action and

application becomes knowledge. Knowledge exists in forms such as instincts, ideas, rules, and

procedures that guide actions and decisions. In our earlier research work we gave a detailed

outline regarding designing of knowledge management system as per organizational need.

When we store the tacit knowledge from the entire knowledge worker naturally the size of

knowledge warehouse goes on increasing day by day. Next step is to access the knowledge as

per our requirement for effective decision making.

All the organizational knowledge is stored in knowledge repository which is separate

from the traditional database. For taking strategic decisions we require this knowledge

warehouse. This knowledge warehouse gives you the knowledge suitable to the present

condition. In this paper we are proposing the application of ant colony optimization technique

to make the optimum selection amongst the provided solutions given by the knowledge

management system so as to improve the process of decision making for the successful

managers.

[2] KNOWLEDGE MANAGEMENT

Knowledge management is based on the idea that an organisation’s most valuable

resource is the knowledge of its people. Therefore, the extent to which an organisation performs

well, will depend, among other things, on how effectively its people can create new knowledge,

share knowledge around the organisation, and use that knowledge to best effect. Knowledge

management is about applying the collective knowledge of the entire workforce to achieve

specific organisational goals. The aim of knowledge management is not necessarily to manage

all knowledge, just the knowledge that is most important to the organisation. It is about

ensuring that people have the knowledge they need, where they need it, when they need it – the

right knowledge, in the right place, at the right time. Knowledge resides in people’s heads and

managing it is not really possible or desirable. What we can do, and what the ideas behind

knowledge management are all about, is to establish an environment in which people are

encouraged to create, learn, share, and use knowledge together for the benefit of the

organisation. [2].

Knowledge management is based on the idea that an organisation’s most valuable

resource is the knowledge of its people. This is not a new idea – organisations have been

managing “human resources” for years. What is new is the focus on knowledge. This focus is

being driven by the accelerated rate of change in today’s organisations and in society as a

whole. Knowledge management recognizes that today nearly all jobs involve “knowledge


August 14


work” and so all staff are “knowledge workers” to some degree or another – meaning that their

job depends more on their knowledge than their manual skills. This means that creating, sharing

and using knowledge are among the most important activities of nearly every person in every

organisation.

Some of the knowledge management definitions are listed below:

“The creation and subsequent management of an environment, which encourages

knowledge to be created, shared, learnt, enhanced, organized and utilized for the benefit of the

organisation and its customers.” - Abell & Oxbrow, tfpl Ltd, 2001

“Knowledge management is a process that emphasizes generating, capturing and

sharing information know how and integrating these into business practices and decision

making for greater organisational benefit.” - Maggie Haines, NHS Acting Director of KM

“The capabilities by which communities within an organisation capture the knowledge

that is critical to them, constantly improve it, and make it available in the most effective manner

to those people who need it, so that they can exploit it creatively to add value as a normal part

of their work.” - BSI’s A Guide to Good Practice in KM

“Knowledge is power, which is why people who had it in the past often tried to make a

secret of it. In post-capitalism, power comes from transmitting information to make it

productive, not from hiding it! -”Peter Drucker

“Knowledge management involves efficiently connecting those who know with those

who need to know and converting personal knowledge into organizational one” – Yankee

Group

Good knowledge management is all about getting the right knowledge, in the right

place, at the right time.

Knowledge in organisations is often classified into two types: explicit and tacit.

[2.1] EXPLICIT KNOWLEDGE

Explicit knowledge is knowledge that can be captured and written down in documents

or databases. Examples of explicit knowledge include instruction manuals, written

procedures, best practices, lessons learned and research findings. Explicit knowledge can be

categorized as either structured or unstructured. Documents, databases, and spreadsheets are

examples of structured knowledge, because the data or information in them is organized in a

particular way for future retrieval. In contrast, e-mails, images, training courses, and audio

and video selections are examples of unstructured knowledge because the information they

contain is not referenced for retrieval.

[2.2] TACIT KNOWLEDGE

Tacit knowledge is the knowledge that people carry in their heads. It is much less

concrete than explicit knowledge. It is more of an “unspoken understanding” about



something, knowledge that is more difficult to write down in a document or a database. An

example might be, knowing how to ride a bicycle – you know how to do it, you can do it

again and again, but could you write down instructions for someone to learn to ride a bicycle?

Tacit knowledge can be difficult to access, as it is often not known to others. In fact, most

people are not aware of the knowledge they themselves possess or of its value to others. Tacit

knowledge is considered more valuable because it provides context for people, places, ideas

and experiences. It generally requires extensive personal contact and trust to share effectively.

Part of the work in developing a loyal, dedicated workforce is establishing recognition

and reward systems to encouraging knowledge worker participation in KM initiatives.

Successful managers recognize that knowledge workers are motivated by a variety of factors,

of which monetary compensation is only one. Even those primarily motivated by money

usually can be encouraged to provide more value to the company by formally recognizing

their contribution to the company’s bottom line. One challenge in recognizing the

contributions of knowledge workers is that their contributions are often intangible. It may be

difficult to quantify relative contributions of intellectual property because metrics are either

inappropriate or subject to interpretation. For example, a programmer who contributes 20,000

lines of code to a project may add less value to the company than one who contributes 2,000

lines of code in one-tenth the time, assuming the code provides the same functionality. [3].

[3] KNOWLEDGE MINING

Improving the quality or accessibility of enterprise data is not an end in and of itself. It

is merely an enabler for creating business value. The data strategy must be driven by an

understanding of how information can enable or improve a business process. For example,

increasing cross-channel sales (a business value) requires data about your current customers

and the products they own (the data); or reducing the cost of manual reconciliation for financial

reporting (the business value) requires standardizing and consolidating redundant and

inconsistent data across business applications (the data).

The data strategy does not need to identify all possible business benefits, but it should

define several that are material to the business and measurable. Establishing some early, visible

benefits is important to launching the data strategy and giving it momentum.

Not all data in the business is critical. In fact, most data is specific to an

application, business function or transaction. Data that is critical typically has two

characteristics:

It is associated with something of long-term value to the firm (e.g., product, customer

or financial information)

• It is used across multiple systems and business processes.

A high-level process flow through marketing, sales, fulfillment and finance for a top

technology company would go something like this:

• Marketing creates interaction information as it reaches out to organizations and individuals

through its marketing campaigns.

• As sales leads emerge, a salesperson is assigned, partners are engaged and sales opportunity

information is maintained throughout the sales process.


August 14


• When an agreement is reached, terms are shared with product fulfillment to deliver the

product and maintain support.

• Finance and sales validate commissions with the sales teams and partners.

• Management uses an end-to-end view of these processes to evaluate the

effectiveness of its pipeline and make ongoing improvements in and across the areas.

This process analysis reveals several critical data assets and associated attributes. For

example, customer organization and individual information is used by every one of the process

steps. If this information is silo and inconsistent, customers will get inconsistent messages and

service. Process owners will have difficulty measuring their effectiveness. Analyses will not

reconcile. And implementing new controls or improvements will require changes within each

process step.

Conversely, improvements to these critical data assets will likely yield business

benefits in all five areas. In our experience, identifying and improving critical data assets in

large companies can yield tens of millions of dollars in benefit, and justify millions of dollars of

investment in implementing a data strategy.

However, we believe it is just as important to keep the set of critical data assets as small

as possible. Note that the most critical data asset for these subject areas is a common identifier.

Maintaining the unique identity of customers, products, interactions and contracts is what links

information across the enterprise. Once that is tackled, attributes can be added incrementally to

the enterprise record over time.

We are proposing to save this critical data in some other repository we call it “

Knowledge Warehouse “ which is managed and updated by knowledge workers. And

extracting knowledge from that repository is “knowledge mining”.

Knowledge mining can thus be characterized as concerned with developing and

integrating wide range of data extracting tools which is capable to extract knowledge from

knowledge warehouse. The designing of physical schema of knowledge warehouse is

customized according to organizations need.

To manage this dynamic environment, the flows of data across systems and processes

need to be organized in a coherent way.

[4] OPTIMIZATION TECHNIQUES

According to definition from business dictionary [1] optimization means finding an

alternative with the most cost effective or highest achievable performance under the

given constraints, by maximizing desired factors and minimizing undesired ones. In

comparison, maximization means trying to attain the highest or maximum result or outcome

without regard to cost or expense. Practice of optimization is restricted by the lack of

full information, and the lack of time to evaluate what information is available.

The different types of optimization techniques are as below:



[4.1] CLASSICAL OPTIMIZATION TECHNIQUE

The classical optimization techniques are useful in finding the optimum solution or

unconstrained maxima or minima of continuous and differentiable functions. These are

analytical methods and make use of differential calculus in locating the optimum solution.

The classical methods have limited scope in practical applications as some of them

involve objective functions which are not continuous and/or differentiable. Yet, the study of

these classical techniques of optimization form a basis for developing most of the numerical

techniques that have evolved into advanced techniques more suitable to today’s practical

problems. These methods assume that the function is differentiable twice with respect to the

design variables and the derivatives are continuous. Three main types of problems can be

handled by the classical optimization techniques:

Single variable functions

Multivariable functions with no constraints,

Multivariable functions with both equality and inequality constraints. In problems with

equality constraints the Lagrange multiplier method can be used. If the problem has

inequality constraints, the Kuhn-Tucker conditions can be used to identify the optimum

solution.

These methods lead to a set of nonlinear simultaneous equations that may be difficult to

solve

[4.2] NUMERICAL METHODS OF OPTIMIZATION

Linear Programming:- Linear programming is the process of taking various linear

inequalities relating to some situation, and finding the "best" value obtainable under

those conditions. It studies the case in which the objective function f is linear and the set

A is specified using only linear equalities and inequalities. (A is the design variable

space).

Integer Programming: - It studies linear programs in which some or all variables are

constrained to take on integer values.

Quadratic programming: - It allows the objective function to have quadratic terms,

while the set A must be specified with linear equalities and inequalities

Non linear Programming :- It studies the general case in which the objective function

or the constraints or both contain nonlinear parts.

Stochastic Programming :- Under this method we study the case in which some of the

constraints depend on random variables.

Dynamic Programming :- Under this type of method we study the case in which the

optimization strategy is based on splitting the problem into smaller sub-problems.

Combinatorial optimization :- This technique is concerned with problems where the set

of feasible solutions is discrete or can be reduced to a discrete one.

Infinite :- dimensional Optimization:-This optimization technique studies the case when

the set of feasible solutions is a subset of an infinite-dimensional space, such as a space

of functions.


August 14


Constraint Satisfaction:- This optimization technique studies the case in which the

objective function f is constant (this is used in artificial intelligence, particularly in

automated reasoning).

[4.3] ADVANCED OPTIMIZATION TECHNIQUES

Hill Climbing: It is a graph search algorithm where the current path is extended with a

successor node which is closer to the solution than the end of the current path. In simple

hill climbing, the first closer node is chosen whereas in steepest ascent hill climbing all

successors are compared and the closest to the solution is chosen. Both forms fail if there

is no closer node. This may happen if there are local maxima in the search space which

are not solutions. Hill climbing is used widely in artificial intelligence fields, for reaching

a goal state from a starting node. Choice of next node/ starting node can be varied to give

a number of related algorithms.

Stimulated Annealing: The name and inspiration come from annealing process in

metallurgy, a technique involving heating and controlled cooling of a material to increase

the size of its crystals and reduce their defects. The heat causes the atoms to become

unstuck from their initial positions (a local minimum of the internal energy) and wander

randomly through states of higher energy; the slow cooling gives them more chances of

finding configurations with lower internal energy than the initial one. In the simulated

annealing method, each point of the search space is compared to a state of some physical

system, and the function to be minimized is interpreted as the internal energy of the

system in that state. Therefore the goal is to bring the system, from an arbitrary initial

state, to a state with the minimum possible energy.

Genetic Algorithm: A genetic algorithm (GA) is a local search technique used to find

approximate solutions to optimization and search problems. Genetic algorithms are a

particular class of evolutionary algorithms that use techniques inspired by evolutionary

biology such as inheritance, mutation, selection, and crossover (also called

recombination). Genetic algorithms are typically implemented as a computer simulation,

in which a population of abstract representations (called chromosomes) of candidate

solutions (called individuals) to an optimization problem evolves toward better solutions.

The evolution starts from a population of completely random individuals and occurs in

generations. In each generation, the fitness of the whole population is evaluated, multiple

individuals are stochastically selected from the current population (based on their

fitness), and modified (mutated or recombined) to form a new population. The new

population is then used in the next iteration of the algorithm.

Ant Colony Optimization: Ant Colony Optimization (ACO) is a meta-heuristic for

solving combinatorial problems. [5, 6, 7, 8]. In ACO, artificial ants construct a solution

by building a path on a construction graph G- (C, α) where the elements of α (called

connections) fully connect C ( set of components). The artificial pheromone can be

associated either to components (nodes) or connections (edges). The behavior of ant is

specified by defining start states and termination conditions, construction rules,

pheromone update rules and domain actions. This has been thoroughly detailed in [5, 6].

In the real world, ants (initially) wander randomly, and upon finding food return to their



colony while laying down pheromone trails. If other ants find such a path, they are likely

not to keep traveling at random, but instead follow the trail laid by earlier ants, returning

and reinforcing it if they eventually find food. Over time, however, the pheromone trail

starts to evaporate, thus reducing its attractive strength. The more time it takes for an ant

to travel down the path and back again, the more time the pheromones have to evaporate.

A short path, by comparison, gets marched over faster, and thus the pheromone density

remains high. Pheromone evaporation has also the advantage of avoiding the

convergence to a locally optimal solution. If there were no evaporation at all, the paths

chosen by the first ants would tend to be excessively attractive to the following ones. In

that case, the exploration of the solution space would be constrained. Thus, when one ant

finds a good (short) path from the colony to a food source, other ants are more likely to

follow that path, and such positive feedback eventually leaves all the ants following a

single path. The idea of the ant colony algorithm is to mimic this behavior with

"simulated ants" walking around the search space representing the problem to be solved.

They have an advantage over simulated annealing and genetic algorithm approaches

when the graph may change dynamically. The ant colony algorithm can be run

continuously and can adapt to changes in real time. This is of interest in network routing

and urban transportation systems.

[5] IMPLEMETATION OF ACO

While working at the strategic level managers tend to take which has direct impact on

the organizations. Still we need human intervention to deal with such type of decision making.

Before coming to any conclusion the various parameters are find out which has a direct impact

on the process. The knowledge related to the process is available through knowledge

management system. We are proposing the use of Ant Colony Optimization [ACO] technique

for the next exploration. What will be the next best strategy or in which manner a particular

problem should be deal can be addressed more efficiently using ACO

A graph G consists of a non empty set of elements called vertices of graph G is called

the vertices set of G, represented by V(G) and the list of edges is called the age list of G denoted

by E(G) . In ACO artificial ants construct a solution by building a path on a construction graph.

Each ant is initially positioned on a randomly chosen node of G and builds a solution by

applying probability rule called as state transition rule. This probabilistic rule is biased by

pheromone value so that higher the pheromone on connection, the higher the profanity that it

will be selected.

Consider that a enterprise want to launch its new product in global market. It has

working in 8 different countries. Now it has to choose whether they global launching or

country wise. Assume that enterprise want to go step wise i.e. country wise launching of their

product; the next job is to find out the optimum route so as to gain the maximum profit and

demographic advantage. Under such situation construct a graph containing countries as their

nodes and level of profitability, resource availability, political environment etc be the

characteristics of the edges connecting the different countries.

Figure 1 illustrates the strategic dependency network. Let us assume that at the time

point F ant K is positioned to find out the best profitable and maximum profitable sequence.


August 14


(We are applying the fuzzy logic [10] approach for deterring the characteristics of variables so

as to get more humanistic approach)

Table 1 illustrates the fuzzy preference assigned to various parameters for edges. The

linguistic terms used are very high, fairly high, high, medium, slightly medium, moderately

poor, fairly poor and poor.

Figure: 1. Enterprise strategic dependency network

Table: 1. Fuzzy linguistic term for selection

In general, an ACO algorithm can be applied to any combinatorial problem as far as it

is possible to define-

1. Appropriate problem representation: The problem must be able to describe as a graph

with as set of nodes and edges between nodes.

2. Heuristic desirability (η) of edges: A suitable heuristic measures of the “goodness” of path

from one node to every other connected node in the graph.

3. Construction of feasible solution: A mechanism must be processed where by possible

solutions are efficiently created.

4. Pheromone updating rule: A suitable method of updating the pheromone levels on edges is

required with a corresponding evaporation rule. Typical methods involve selecting the n best

ants and updating the paths they choose.

5. Probabilistic transition rule: The rule that determine the probability of an ant traversing

from one node in the graph to the next. The features selection task may be reformulated in to an

ACO – suitable problem. ACO requires a problem to be represented as a graph where nodes

represent features with the edges between them denoting the choice of the next future. The



search for the optimal feature subset is then an ant traversal through the graph where a

minimum member of bodes are visited that satisfies the traversal stopping criterion.

Figure: 2. Illustration of the setup

The ant is currently at node F and has a choice of which to select next so as to add to its

next path i.e. dotted lines. It chooses country D next based on the transition rule and in the same

manner E, B, E, A, C,H and then G. upon arrival on G the current set is determined to find the

optimum sequence for perfect launching of the product. The heuristic desirability of traversal

and edge pheromone levels are combined to form the so- called probabilistic transition.

Depending on how optimality is defined for the particular application, the pheromone is

updated accordingly. For instance, here the goodness of the pheromone is directly proportional

to profitability, economy and inversely proportional to political benefit. There is also the

possibility of avoiding any country from the launching process.

The process begins by generating a number of ants K, which are then placed randomly

on the graph i.e. each ant starts with on random country. Alternatively, the numbers of ants to

place on the graph may be set equal to the number of features within the data; each ant starts

path construction at a different country. From these initial positions, they traverse edges

probabilistically until a traversal stopping criterion is satisfied. The resulting subsets are

gathered and then evaluated. If an optimal subset has been found then the best route subset

encountered. If neither condition holds, then the pheromone is updated, a new set of ants are

created and the process iterates once more.

[6] CONCLUSION

Taking strategic decisions are most crucial activity for the existence of any enterprise in

a competitive business model. The decisions involved in these process are extremely uncertain

and require plenty of strategic interventions. In this paper we formulate the use of ant colony

optimization technique along with fuzzy linguistic variable so as to give more humanistic

approach for the decision making process. The work is explained by giving a example of

product launching at eight different countries.


August 14


REFERENCES

[1] http://www.businessdictionary.com/definition/optimization.html#ixzz3BnMVtrfL

[2] Bergeron Bryan, Essentials of Knowledge Mnagement, Wiley, 2003, pp 90

[3] Chen Hsinchun , Knowledge Management System A Text Mining Perspective , 2001 pp 8-

9

[4] Brun Caroline De, NHS National Health Library for health, July 2005, pp3-6.

[5] Dorigo m and Caro G Di, ” The Ant Colony Optimization Meta- Heuristics”, New Ideas in

Optimization, McGraw – Hill, London, pp 11-32

[6] Dorigo M. And L M Gambardella, “Ant Colony System : A cooperative learning approach

to the travelling salesman problem”, IEEE transactions , Evolutionary Computation, 1997,

pp53-66

[7] Dorigo M, Maniezzo V and Colony A, “Ant Colony System: Optimization by a Colony of

Cooperating Agents”, IEEE Transactions System Man and Cybernetics, Part B, 26 (1)

1996 pp 29-41

[8] Gary M R , and Johnson D S, “ Computers and Intractability : A guide to the theory of NP

Completeness”, W H Freeman 1979

[9] Lokketangen A, Satisfied Ants, in ANTS 2000, 2000

[10] Zadeh L. A., “Roles of Soft Computing and Fuzzy Logic in the Conception”, Design and

Deployment of Information/ Intelligent Systems, Computational Intelligence: Soft

computing and Fuzzy- Neuro integration with Applications, O Kaynak, L A Zadeh, B

Turksen, I J Rudas, 1998, pp 1-9.

AN APPLICATION OF ANT COLONY OPTIMIZATION IN A KNOWLEDGE MINING ENVIRONMENT

Documents

Transcript of AN APPLICATION OF ANT COLONY OPTIMIZATION IN A KNOWLEDGE MINING ENVIRONMENT