The History Of Adaptive Business Intelligence

Published Date: 02 Nov 2017

Dr. Iman Ardekani

Class:

3003

Deadline:

20th April,2013

Subject:

ISCG-8043 Adaptive Business Intelligence

File Name : Adaptive Business Intelligence Assignment 1

ISCG-8043

Adaptive Business Intelligence assignment 1

Abstract

This report contains the study of Adaptive business intelligence. As the world is at developing stage in telecommunication, Adaptive business inelegance is also gaining importance day by day. Many business sectors follow Adaptive business intelligent approach to study the detailed customer behavior for obtaining maximum profit and minimum loss. Data mining one of the important Adaptive business techniques are reviewed in this report. Basic data mining algorithms and data structures are explained in the report. This report contains the business sectors who make use of adaptive business intelligence for the marketing of their products.

Contents

Introduction

Many a times this intelligence is identified by a kind of behavior that allows an individual to change a disruptive behavior to something more advanced behavior to something more constructive. These behaviors are many time social or personal behaviors. Thus a behavior can b adapted to something else. A human being or an animal makes decisions continuously without any central control or coordination. Since the computer age unfolds on humanity, one of the most important areas in information technology has been that of decision support. In today’s real world this area is most important than ever. Today working in dynamic and ever-changing environment, modern-day managers are back bone for an assortment of effective decisions.

Adaptive business intelligence brings forward some of the important questions such as:

The workforce of the company increased or decreased?

Start up with new market interest?

Developing new products?

Investing in research and development?

To create an effective adaptive business intelligence system it should consist of three major components:

The element of making predictions such as sales price prediction.

The element for making near-optimal decisions like the proper distribution of a product.

A component for adapting the prediction module to changes in the environment.

Hence we can say Adaptive business intelligence is a set of theories, methodologies, processes and architectures for business purposes. It can handle large ammont of information to help identify send develop new opportunities. For long-term stability and market advantages, new opportunities and implementation of effective strategy is made use of. The functions of adaptive business intelegence are same as that of competitive intelligence in today’s world. Because both the ideas work on decision making.

Self-adaptive software is one the most important discoveries in the field of Adaptive business intelligence. Originally discovered as evolution strategies in Germany. Due to its Self-adaptive nature it enables the algorithm to dynamically adapt to the problem characteristics and even to cope with ever changing environmental conditions as they occur in unforeseeable ways in many real-world business applications.

Over the past 50 years, computer science has seen many number of fundamental inventions and, coming along with them, revolutions such as how software systems are able to deal with data.

50es. Von Neumann architecture, simple operating systems, most basic programming languages.

60es. Improved operating systems (especially UNIX), structured programming, object-oriented programming, functional and logic programming.

70es. Relational model of Codd, relational database management systems (RDBMS).

80es. Enterprise resource planning (ERP) systems, production planning (PPS) systems, reflecting an integrated toolbox on top of the RDBMS-level.

90es. World wide web and internet-programming, facilitating a world-wide integrating access to ERP- and PPS-systems.

Adaptive business intelligence

The methodology focuses completely on the business relevant aspects, such as the business input and output. Business input means the problem to be solved, together with the corresponding data, and business output is the problem knowledge or problem solution generated by the approach. The important tasks on the way from data and problem to an optimal business solution are data mining, data analysis and optimization tasks. For performing this tasks, computational intelligence today offers technologies which allows us to solve complex problems which was not possible in early days.

img 1.jpg

Figure1: Adaptive business intelligence model

In data analysis and data mining, the task is to find out hidden knowledge from large amount of data. Like in financial applications, chemical processes, marketing data and many other fields. The term knowledge simplifies that the output should be compact, presented in a symbolic way, interpretable and highly relevant to the data. Derived knowledge and mathematical model are also used together so that the knowledge is incorporated into a model. Technologies from computational intelligence which support various aspects of data mining process includes classifier systems, genetic programming, fuzzy logic, and rough sets.

Data mining

Data mining is known as data or knowledge discovery. It is the process of studying data from different prospective and summarizing it into use full information, such research can help us to increase the output and decrease the cost and time spend on it. Data mining allows users to analyze the data from many different areas and categorize it, and summarize the relationships identified. Technically, data mining is the process of obtaining correlations of patterns among many different fields in large relational database. Data mining majorly focuses on retail business, financial business, communication sector, and marketing sector organizations. It helps these companies to find out relationships among "internal" factors such as price, product positioning, or staff skills, and "external" factors such as economic indicators, competition, and customer demographics. And, it allows them to find out the impact on sales, customer satisfaction, and corporate profits. Finally, it obtains them to search deep into summary information to view detail transactional data.

Untitled.jpg

Figure 2: Database

Modern science and engineering are based on using first-principle models to describe physical, biological and social systems. This approach starts with a scientific model, such as Newton’s law of motion or Maxwell’s equations in electromagnetism, and then builds upon them various applications in mechanical engineering or electrical engineering. As data mining is a natural activity to be performed on large data sets, one of the largest target markets is the entire data-warehousing, data-mart, and decision-support community from the industries such as retail industries, manufacturing industries, telecommunication sector, health care department, insurance sector, and transportation department. In the business community, data mining technology can b used to discover new purchasing patterns, plan investment strategies, and detect unauthorized expenditures in the accounting system. Data mining techniques can be applied to problems of business process reengineering, in which the goal is to understand interactions and relationships among business practices and organizations.

In simple words data mining is referred to extracting or mining knowledge from large amount of data. Mining is an intense term characterizing the process that finds a small set of precious nuggets from a great deal of raw material. Thus, such a misnomer that carries both data and mining became a popular choice. Many people treat data mining, as a synonym for another popularly used term, knowledge Discovery from Data, or KDD.

s.jpg

Figure 3: Data mining process

Data mining process:

Data cleaning – the first step in data processing is data cleaning, it involves removing of noise full and inconsistent data from the datasets.

Data integration – in this step multiple data sources are combined for obtaining desired output.

Data selection – in this step relevant or useful data is retrieved from the database.

Data transformation – in this step data are transformed or consolidated into forms appropriate for mining by performing summary or aggregation operations.

Data mining – this step is an essential process where intelligent methods are applied in order to find out data patterns.

Pattern evaluation – this step helps to identify the truly interesting patterns representing knowledge based on some interesting measures.

Knowledge presentation – this is the final step in data mining process, where visualization and knowledge representation techniques are used to present the mained knowledge to the user.

Knowledge discovery database (KDD)

Knowledge discovery database is a process of finding out useful information and patterns in data.

Example: Web Log

Select the log data which consists of dates and location to use.

Remove identifying URLs and Error logs.

Transformation or sorting and grouping.

Identifying and counting patterns, constructing a data structure.

Identifying and displaying frequently accessed sequences of data.

And the final step of cache prediction and personalization.

Data mining Algorithm

A data mining algorithm is a set of calculations that creates a data mining model from data. While creating a model, the algorithm first analyzes the presented data, searching for specific types of patterns or trends (similar kind of data). The algorithm uses the results of this analysis to define the optimal parameters are then applied across the entire data set to extract actionable patterns and detailed statistics.

Data mining methods

Predictive modeling also knows as classification or regression methods.

Classification and regression assign a new data record to one of several predefined categories or classes. Regression deals with predicting real-valued fields, it is also called as supervised learning.

Segmentation also known as clustering

Partition the dataset into subsets or groups such that elements of a group share a common set of properties, with high within group similarity and small inter-group similarity, it is also called as unsupervised learning.

Summarization also known as associations

Association detects sets of attributes that frequently co-occur, and rules among them. For example: 90% of the people who buy bread, also buy milk (60% of all grocery shoppers buy both)

Sequence mining

It discovers sequences of events that commonly occur together. E.g.: In a set of DNA sequence ACGTC if followed by GTCA after a gap of 9, with 30% probability.

model

Figure 4: Data mining methods

Simple data mining example

Find out all credit applicants with last name Smith? This is an example of Classification.

Identify customers who have purchased more than 10,000$ in last four months? This is an example of clustering.

Find all customers who have purchased milk? This is an example for association rule.

Adapting Data mining technology in business

The economics of customer relationship are changing in fundamental ways, and companies are facing the need to implement new solutions and strategies that address these changes. The concept of mass production and mass marketing, first started during the industrial revolution. The tools and technologies of data warehousing, data mining and customer relationship management (CRM) afford new opportunities for business to act on the concept of relationship marketing.

There are number of industries that are using data mining applications. Some of these organization include retail stores, hospitals, banks, insurance companies, manufacturing combining with data mining techniques like statistics, pattern recognition and other important tools used to find patterns and combinations which are difficult to find in real world.

Decision trees

A decision or a classification tree represents a set of rules that can be explained in the form of spoken language and can be easily converted to programming language like SQL. Private health insurance most commonly make use of decision tree. Health Insurance Company follows a strict rule for accepting or rejecting a application. If a application is rejected than the company has to provide an adequate reasons to avoid claims of discrimination.

Types of regression

Regression is a way of association between dependent variable and one or more independent variables. This associations are formulated as equation, where independent variable have parametric coefficient that allows future values of the depend variable to be predicted.

The two main types of Regression methods are:

Linear regression: in linear regression the dependent variable is continuous and in logistic form it is either discrete or categorical.

Logistic regression: in logistic regression the discrete variable must be transformed into a continuous value that is a function of the probability of the event going on.

Neural networks

Traditionally the term neural network was used for defining network or circuit of biological neurons. But now a day’s neural network terms is used for explaining artificial neural networks, which consist of set of nodes. There are two types of neural network models: supervised neural network and unsupervised neural network.

Retail Marketing

The use of store-branded credit cards and point-of-sale systems, retailers can keep record of every shopping transactions. This enables to better understand various customer segments. Some retail applications include performing basket analysis also known as affinity analysis, basket analysis reveals which product is costumer very often purchase. This knowledge can improve stocking, store layout strategies and promotional offers.

Sale forecasting – examining time based patterns helps retailer make adjust available stock. If a customer purchases an item, then what’s the time prediction for purchasing the complementary item for that product.

Database marketing – retailers can create a customer profile depending upon the certain behaviors of the customers, like those peoples who purchase designer labels clothing or those people who attend sales. This information can be helpful for effective cost promotion.

Merchandise planning and allocation – when retailers opens a new stores, they can improve merchandise planning by examining patterns in store with similar characteristics. Retailer can also use data mining to find out the ideal layout for a specific store.

Banking sector

Banks can utilize knowledge discovery for various applications such as,

Card marketing – by identifying customer segments, card applicants and acquires can improve profitability with more effective accretion and confined programs, targeted product development and customized pricing.

Cardholder pricing and profitability – card applicants can make use of data mining technology to price their products so as to maximize profit and minimize loss of customers.

Fraud detection – fraud is enormously costly. By analyzing past transaction that were later determined to be fraud can help bank in finding out the transaction pattern.

Predictive life-cycle management – data mining helps banks predict each customer’s importance and to service each segment appropriately like offering special deals and offers.

Telecommunication sector

Telecommunication companies around the world face escalating competition which is forcing them to aggressively market special pricing programs aimed at retaining exiting customers and attracting new ones.

Call detail record analysis – telecommunication companies acquire detailed call records. By identifying customer segments with similar use patterns, the companies can design attractive pricing and feature promotions.

Customer loyalty – some customers again and again switch providers, or agitate to take advantage of attractive incentive by competing companies. The companies can use data mining for studying the characteristics of customer who are likely to remain connected to the current network providing company; this allows the companies to target their spending on customers who will be more profitable to them.

Insurance companies

Insurance companies are facing problems of mailing costs, increasing marketing campaigns, cross selling to existing customers.

Results: effectiveness of its campaigns, optimization and execution, decreased mailing costs and increase conversion rates.

Future of Adaptive business Intelligence

In today’s developing world, analyzing data to predict market interest and to improve company performance is an very essential business activity. However, its becoming clear that business success requires very careful data analysis to be performed in real-time in order to meet the rapidly changing demands from both customers and regulators. Due to adaptive business intelligence techniques machine intelligence can be defined as the system to adapt its behavior to meet desired output in different environments. Captivating the tree components of prediction, adaptation and optimization becomes the back bone for core module of adaptive business intelligence system. Definitely the future of the adaptive business intelligence industry lies in organizations that can make decisions, rather than making use of detailed reports producing tools. Adaptive business intelligence had also provided the way for advancing the technologies of artificial intelligence like robotic and automation on space station. This technology will be very beneficial to the economic development of the country.

Conclusion

Adaptive business intelligence is one of the fastest growing studies for business development. It helps the company for maximizing the profit and minimizing the loss. It’s the way of designing a sale or service pattern depending upon the critical interest of a customer.

The best tool for using data mining is the question, "what is available and what gives the best results". Neural networks tool is one of the best tool for data mining operation also gives very promising results.