Big Data: Map Reduce Based Outlier Identification for Fraud

Big Data: Map Reduce Based Outlier Identification for Fraud Big Data: Fast, Parallel Map Reduce based outlier identification for Fraud and Intrusion Detection Pooja Vijay Pawar Ann Marie Joy Abstract One of the most challenging aspect of Big Data analytics is real time monitoring of data. Larger and larger amount of data is being collected and stored on a daily basis, thus increasing the need for quick, effective and efficient way of analysing the data in determining potential malicious data. Also network security threats are increasing at an alarming rate and are becoming increasingly sophisticated and difficult to detect. Web traffic generated by non-human activities such as botnets or worms consume network resources, delude people and affect network security. Most of the existing work in Fraud Detection/Intrusion Detection regard being an outlier as a binary property. In this paper we use a relative density based approach implemented as a MapReduce job, which gives a sense of degree of a data point being an outlier; this is more meaningful and also better immune to false positives. Keywords Big Data, Fraud detection, Intrusion Detection, Hadoop, Outlier, Cluster, Security, Relative Density, LOF 1. Introduction We are currently living in a world where we are surrounded and ruled by data. Continuously, exponentially huge amounts of data is collected, stored, processed and made available in a variety of forms every day. Recently, Network Intrusion and Fraud detection has received increased attention with regard to network security, mainly due to this. Big Data is drastically changing the way in which we detect fraud and intrusion in real time using advanced analytic solutions that are very powerful, complex and fast. In this paper, we propose a methodology to detect Fraudulent Web traffic and Intrusion in a network using MapReduce-based outlier detection. These features help in filtering out clients that generate abnormal traffic and specifically show different levels of potential anomalous traffic for each suspicious client. The detected abnormal web traffic can be visualized easily and this method can be implemented for large networks and can be scaled accordingly. Outlier can be defined as a data which is very distinct from the other data of the same dataset, based on some distance measure. Outlier detection, being a significant data mining problem has engendered a lot of research interest in the recent past. As a result, various methods for outlier detection have been developed particularly for dealing with numerical data. However, outlier detection for categorical data still remains an unexplored field. Addressing this requirement, we propose a two-phase algorithm for detecting outliers in categorical data based on a novel definition of outliers. This algorithm initially explores a clustering of the given data which is followed by the ranking phase for determining the set of most likely outliers. The proposed methodology is expected to show better results as it can identify different types of outliers, using independent ranking scheme based on the inherent clustering structure of the given data. Hadoop is a very popular open source Apache project, which is used for storing and processing huge volume of data on commodity hardware. Hadoop package mainly consists of MapReduce Engine and Hadoop File system. Many Frameworks have been built on top of Hadoop. Using the distributed architecture of Hadoop in this paper we discuss how we can exploit it for identifying outliers in captured network data. 2. Related Work There has been lot of research and number of techniques which have been developed in the past two decades with respect to fraud detection and Intrusion Detection. Lot of machine learning techniques such as Neural Networks, Markov model, K Nearest Neighbour have drawn special attention. In this paper we use unsupervised Machine Learning techniques to identify fraudulent transactions using Hadoop. Most current practices to the process of detecting intrusions exploit some form of rule-based study. Rule-based analysis depend on sets of predefined rules that are configured by an administrator and are automatically created by the system. The use of automated system techniques in intrusion detection mechanisms was a significant milestone in the development of effective and practical detection based information security systems. Rule-based systems suffer from an incompetence to detect attacks situations that may occur over an extended duration of time. However most significant benefit of neural networks in intrusion detection is the capability of the neural network to learn the characteristics of abnormal attacks and recognize instances that are unlike any which have been detected before by the network. Majority of studies that proposed Hidden Markov Model to implement IDS are related to host-based systems, i.e., IDS that analyses the actions performed on a single host to detect attempts of intrusion. 3. Outlier Detection Algorithms Outliers can be detected using various different techniques. Some of the techniques are discussed below 3.1 Distance-based and Clustering Approaches Distance-based methods do not make conventions for the data since they basically compute the distance between each point. For example, Knorr et al. Proposed a k-NN technique where, if m of the k nearest neighbours of a point are within a specific distance d, then the point can be classified as normal. Knorr et al. points as an outlier if at least p% of the points in the dataset lie more than distance 10 d from it. These methods show high computational complexity (e.g. nearest neighbour based methods have quadratic complexity with respect to the number of data points) which renders them impractical for really large datasets. Several approaches may be engaged to make the k-NN queries faster (to achieve linear or logarithmic time), such as an indexing structure (e.g. KD-tree, or X-tree); however these structures have been shown to break down as the dimensionality grows. Clustering can be defined as the task of grouping a set of objects in such a way that objects in the same group share some feature similarity among each other than to those in other groups. Clustering is one of the very popular techniques currently being used in outlier detection. Any object that has weak membership to the cluster it belongs to is a potential outlier. If there are any small clusters from other clusters, then the smaller cluster could potentially be an outlier. For instance there could be many different kinds of fraudulent transactions which might have high similarity among themselves and form a cluster. There are additional problems with clustering. Clustering algorithms are optimized to find clusters rather than outliers. Hence, sometimes it may be hard to tell whether a cluster belongs to fraudulent transactions or some new emergent buying behaviour of a legitimate user. Hence before making a final call we must perform additional analysis. 3.2 Statistical distribution Statistical outlier detection was one of the most basic approaches dating back to the 19th century. Multivariate statistical methods have been proposed, together with use of robust outlier’s estimates of the multidimensional distribution parameters, e.g. minimum covariance 9 determinant (MCD) and minimum volume ellipsoid (MVE). One critical problem of statistical-based methods is the suitable model for each dataset and application. Also, as data rises in dimensionality, it becomes ever more perplexing to estimate the multidimensional distribution. As the data increases in dimensionality, data ranges through a larger volume and becomes sparse. In addition to the reduction this has on performance, it also spreads the convex hull, thus altering the data distribution. This can be improved by pre selecting the most noteworthy features to work, projecting to a lower-dimensional subspace, or applying Principal Component Analysis (PCA). Another methodology to deal with higher dimensio nalities is to organize data points in convex hull layers according to their peeling depth, based on the idea that outliers are data objects with a shallow depth value. Fig 1 In this approach one main assumption is that data objects follows a certain distribution (E.g. Gaussian) and normal data objects occur in a high probability region of this model. Fig.1 shows an example where there is high concentration of data points lying in the normal region which associates to normal data points; the mini distributions on both sides of the normal distributions are possible outliers. As shown in Fig.1 outliers will deviate strongly from this distribution. There are lot of issues with this technique too, main being the curse of dimensionality, other being lack of robustness. This is because Mean and standard deviation are very sensitive to outliers. 3.3 Density Based approach In this technique we compare the density around a data point with the density around its local neighbours. The computed density is called an outlier score. The main assumption here is that the density around a normal data point is almost similar to the density around its local neighbours. Here Density/Outlier score means that some clusters are densely packed and some others are not. Mathematically, it is defined as the inverse of the average distance to the k nearest neighbours. Lower density of a data point signifies that the probability of it being an outlier is very high. There have been many variants of Density based approaches suggested in the past few decades, majority of which deal with decreasing the computational time complexity. In Fig.2 the points which are densely packed, appearing yellow indicate normal data points, the ones which are away from the cluster are outliers, possible candidates for malicious data. In this paper, we use this technique to determine possible candidates for outliers. 4. Experiment For our experiment we used KDD Cup 1999: Computer network intrusion detection dataset for testing and evaluating our approach. We used Relative Density [3] based approach for our system. Which involved 4 Map Reduce Tasks. The Algorithms works as follows: Computing K-NN [3] We begin with the notion of K-Nearest Neighbour of object p. Definition: K-Nearest Neighbour of object p For any positive integer k, the k-distance of object p, denoted as k-distance (p), is defined as the distance d(p,o) between p and an object o ∈ D such that: (i) For at least k objects o’∈D{p} it holds that d(p,o’) ≠¤ d(p,o), and (ii) For at most k-1 objects o’∈D{p} it holds that d(p,o’) . Given the k-distance of p, the k-distance neighbourhood of p contains every object whose distance from p is not greater than the k-distance, i.e. Nk-distance (p) (p) = { q ∈ D{p} | d(p, q) ≠¤ k- distance(p) } These objects q are called the k-nearest neighbours of p. We use 2 MapReduce tasks, one to compute the pairwise distance between data points as explained above and other to compute the density of the data point. The density of a data point is simply the inverse of the average distance to the k nearest neighbours. Finding all the neighbourhood group the data points are associated with and also give them unique id We define one more term reachability distance of an object p w.r.t the data point o to determine the neighbourhood. Definition: Reachability distance of an object p w.r.t. object or Let k be a natural number. The reachability distance of object p with respect to object o is defined as reach-distk(p, o) = max { k-distance(o), d(p, o) }. The higher the value of k, the more similar the reachability distances for objects within the same neighbourhood. We use the same MapReduce class as before with slightly different configuration to identify the neighbourhood. Once neighbourhood are identified they are given a unique ID. Using previous results, create a mapping between data point and its density. In a typical density-based clustering algorithm, there are two parameters that define the notion of density: (i) a parameter MinPts specifying a minimum number of objects; (ii) a parameter specifying a volume. These two parameters determine a density threshold for the clustering algorithms to operate. That is, objects or regions are connected if their neighbourhood densities exceed the given density threshold. To detect density based outliers, however, it is necessary to compare the densities of different sets of objects, which means that we have to determine the density of sets of objects dynamically. Therefore, we keep MinPts as the only parameter and use the values reach-distMinPts(p, o), for o ∈ NMinPts(p), as a measure of the volume to determine the density in the neighbourhood of an object p. Definition: Density of an object p lrdMinPts(p) = 1/ Intuitively, the local reachability density of an object p is the inverse of the average reachability distance based on the MinPts- nearest neighbours of p. Note that the local density can be ∞ if all the reachability distances in the summation are 0. This may occur for an object p if there are at least MinPts objects, different from p, but sharing the same spatial coordinates, i.e. if there are at least MinPts duplicates of p in the dataset. For simplicity, we will not handle this case explicitly but simply assume that there are no duplicates. Hence in our MapReduce implementation, first we sort the data points based on density data and the neighbourhood, such that in the input for the reducer, we get first value as density, and the subsequent values are the neighbourhood ids. Determining the Relative Density or LOF (Local Outlier Factor) Results from the previous step is then used in another MapReduce task to compute the relative density or also called as Local Outlier Factor (LOF). Definition: LOFMinPts(p) = The outlier factor of object p captures the degree to which we call p an outlier. It is the average of the ratio of the local reachability density of p and those of p’s MinPts-nearest neighbours. It is easy to see that the lower ps local reachability density is, and the higher the local reachability densities of ps MinPts-nearest neighbours are, the higher is the LOF value of p. In the following section, the formal properties of LOF are made precise. To simplify notation, we drop the subscript MinPts from reach-dist, lrd and LOF, if no confusion arises. Finally, Data Points which have low relative Density or LOF are determined as possible candidates for outliers. 5. Conclusion Existing Intrusion detection system are in nascent stage in handling extremely large traffic and the data transfers in large Networks. MapReduce Framework can handle large amount of data quickly and efficiently. Thus our proposed methodology for Outlier detection using Relative Density based approach not only can handle large amount of data but also scales easily. In near future full of MapReduce based IDS needs to developed and evaluated. We also plan to explore multiple classifier system compared to single classifier to get improved results. 6. Acknowledgement This work is supported by CSE Department, PES Institute of Technology. 7. References [1] Barnett, V., Lewis, T. (1995). Outliers in Statistical Data. Wiley, 3rd Edition. [2] Davide Ariu, Giorgio Giacinto, and Roberto Perdisci, Sensing attacks in Computers Networks with Hidden Markov Models. [3] Ng, Jorg Sander, Hans-Peter Kriegel, Raymond T, Markus M. Breunig. LOF: Identifying Density-Based Local Outliers [4] Manh Cong Tran, Lee Hee Jeong, Yasuhiro Nakamura. Abnormal Web Traffic Detection Using Connection Graph. [5] Suri, N.N.R.R. Centre for AI Robot., Bangalore, India Murty, M.N. ; Athithan, G..An algorithm for mining outliers in categorical data through ranking. [6] Kuo Zhao, Liang Hu. Intrusion Detection and Prevention in high speed network.[7] Qing He Yunlong Ma, Qun Wang, Fuzhen Zhuang, Zhongzhi Shi.Parallel Outlier Detection Using KD-Tree Based on MapReduce [8] Koufakou, A. Sch, FL Secretan, J., Reeder, J., Cardona, K., Georgiopoulos, M.Fast parallel outlier detection for categorical datasets using MapReduce. [9] Ganesh Ananthanarayanan, Srikanth Kandula, Albe rt Greenberg, Ion Stoica, Yi Lu, Bikas Saha, Edward Harris . Reining in the Outliers in Map-Reduce Clusters using Mantri. [10] H. GuÃÅ'ˆnes KayacÄ ±k, A. Nur Zincir-Heywood, Malcolm I. Heywood. Selecting Features for Intrusion Detection:A Feature Relevance Analysis on KDD 99 Intrusion Detection Datasets. [12] E. Eskin, A. Arnold, M. Prerau, L. Portnoy, S. Stolfo, â€Å"A geometric framework for unsupervised anomaly detection: Detecting intrusions in unlabeled data,† in Applications of Data Mining in Computer Security, Chapter 4, D. Barbara and S. Jajodia (editors), Kluwer. [13] Q. He, F.Z. Zhuang, J.C. Li, Z.Z. Shi. Parallel implementation of classification algorithms based on mapreduce. International Conference on Rough Set and Knowledge Technology. [15]Koufakou, A., Ortiz, E., Georgiopoulos, M., Anagnostopoulos, G., Reynolds, K., A Scalable and Efficient Outlier Detection Strategy for Categorical Data, Intl Conference on Tools with Artificial Intelligence ICTAI, October, 2007. [16] Big Data Analytics for Security Intelligence, CLOUD SECURITY ALLIANCE September 2013. [17] DuMouchel W., Schonlau M.: â€Å"A Fast Computer Intrusion Detection Algorithm based on Hypothesis Testing of Command Transition Probabilities†, Proc. 4th Int. Conf. on Knowledge Discovery and Data Mining, New York, NY, AAAI Press, 1998, pp. 189-193. [18] Ramaswamy S., Rastogi R., Kyuseok S.: â€Å"Efficient Algorithms for Mining Outliers from Large Data Sets†, Proc. ACM SIDMOD Int. Conf. on Management of Data, 2000. [19] Fawcett T., Provost F.: â€Å"Adaptive Fraud Detection†, Data Mining and Knowledge Discovery Journal, Kluwer Academic Publishers, Vol. 1, No. 3, 1997, pp. 291-316. [20] Holtz, Marcelo D., Bernardo M. David, and Rafael Timà ³teo de Sousa Jà ºnior. Building Scalable Distributed Intrusion Detection Systems Based on the MapReduce Framework. Telecomunicacoes (Santa Rita do Sapucai) 13 (2011): 22-31. [21] DuMouchel W., Schonlau M.:  "A Fast Computer Intrusion Detection Algorithm based on Hypothesis Testing of Command Transition Probabilities†, Proc. 4th Int. Conf. on Knowledge Discovery and Data Mining, New York, NY, AAAI Press, 1998, pp. 189-193. 1 Onida Electronics: New Product Launch Onida Electronics: New Product Launch LAUNCHING OF A NEW PRODUCT ONIDA LAPTOPS COMPANY DESCRIPTION ONIDA is one of the largest television manufacturing companies in INDIA. It is one of the largest and rapid growing companies in the same field. Onida as a company was founded in 1981 as a public company. Its head quarter is in MUMBAI, MAHARASHTRA, INDIA. Onida deals in a type of industry known as electronic industry. Onida Company started their business with electronics and then expanded their business with various products as LCD TVS, PLASMA TVS, TELEVISIONS, DVD AND HOME THREATER SYSTEM, AIR CONDITIONERS, WASHING MACHINES, MICROWAVES, PRODUCTS RELATED TO PRESENTATION, and INVERTERS AND ALSO MOBILE PHONES. Onida Company is a most popular brand now. It has got his network as 33 branch offices and 208 customer relation centers. Also the company is having 41 depots spread all over the country. Onida is also having a market capitalization of 400 crore approx. also onida electronics won an award as AWARD FOR EXCELLENCE IN ELECTRONICS in 1999 from MINISTRY OF INFORMATION TECHNOLOGY, GOVERNMENT OF INDIA. EXECUTIVE SUMMARY Onida is one of the largest rapid growing companies. It is one of the most successful companies in term of professionalism. The main objective is to prove quality of competitive prices and customer satisfaction. It is one of the objective is to provide NOTEBOOK PC as ensured quality, availability and customer satisfaction. This following marketing plan forms the basis for the launching of ONIDA LAPTOPS by ONIDA COMPANY one of the famous company in India. The analysis will allow us to follow for the achievement of the companys strategic goals. ONIDA LAPTOPS will be marketed to reinforce the companys status as leader in innovation and successful product launching. This new product launching will enable us to add RS.100 crores to companys turnover with a forecasted sales growth prospect of 10-15% over the next 5 years. And while satisfying the need and designing the product to match individuals lifestyle. Success will be reflected by a sizeable capture of market shares within this market, while strategically carrying the company up to the top spot as the market leader in laptop segment. Export potentials in the market will be considered in all the western countries as USA, Russia, France, and Iraq. Nigeria, Yemen and even further expansion can be planed. The main aim of the company is to attract a sizeable market share of the laptop segment. Also the number of customers depending upon laptops rather than personal computers is taken care off. The main aim is to sell one lakh units of laptops in the first year considering as this brand as a new brand and being expertise in the related field having manufactured the product for several brands and having also sold its own televisions. EXPECTED TURNOVER Above chart shows the expected turnovers in crores over the coming years in turnover of onida laptops in various countries. In 2009 as the product is being launching in 2009 only so started expected turnover will be near about 90 lakhs. With the continuous supply of good quality and good work with good customer service expected turnover will start increasing. Company is planning to use new technology and other servicer beneficial for customers and company in terms of turnover which will help in increase the expected turnover to 150 lakhs in 2012. This proves that launching of onida laptops will be beneficial for the organization. SITUATIONAL ANALYSIS ONIDA NOTEBOOK PC is a new expansion line of onida. Onida television portfolio has been well received and now ONIDA NOTEBOOK PC is going to be launched for onidas continues success and future profitability. Onida as a company is famous for quality products at reasonable prices and it offer best facilities. It will offer NOTEBOOK PC or LAPTOPS in various configuration as well as prices to satisfy every kind of customer. One important key to success or for the development of the product is to create product awareness and growth of customer base. MARKETING SUMMARY Marketing summary of onida laptops possess sufficient information about the market needs and customer needs, wants, desires and demands. It will also help to give right offer to right customer so that company can achieve customer satisfaction and can communicate with them in a better way. Also this lead to have some important information about customer which will help to make customer a delightful customer. TARGET MARKET The target market of ONIDA LAPTOPS or NOTEBOOK PC is shown below: Above shows that there is a target market of onida laptops as CORPORATE USERS END USERS OTHERS Environmental analysis Macro environmental factors Growing disposable income and low penetration levels would ensure greater share of wallet for the consumer industry and would help in achieving the predetermined targets. The inventions and innovations would help in shaping the industrys future. It would be necessary to catch the changing trends in consumer lifestyles and offer the right product at right time to facilitate further growth and achievement of targets. There is a need to launch a product which will give maximum customer satisfaction and satisfy there needs. Micro environmental factors The cost would further come down with enhanced possibilities of better and easy cross border talks with suppliers worldwide and the range of offerings Would go up. With the increase in number of players, there would be more action and the industry would feel the buzz around it and would need constant high pitch communication with the customers. Network has a strong influence on the buying decision and the organization that would manage the channel better, would have a substantial edge over the others. Competitive Strategy Onida will try to offer more valuable offerings to the customers and would try to manage customer relationships. Onida as a company in order to be effective and ahead of competition would try to maximize benefits to the consumers. They will also react strongly to price changes made by the competitors and organizations would look at integration on backend as well as the front ends. MARKET DEMOGRAPHICS The profile of ONDA NOTEBOOK PC or LAPTOPS customer consists of the following GEOGRAPHIES, DEMOGRAPHICS, and BEHAVIOR FACTOR: GEOGRAPHICAL FACTORS Geographical factors have been classified as: ONIDA NOTEBOOK will have specific domestic geographic target area. They will serve the product to domestic market. Onida will try to cover the Metropolitan area through their own distribution channel. DEMOGRAPHICAL FACTORS Demographical factors have been classified on the basis of the following attributes: There will be almost same featured notebook pc or laptop for the corporate, end users and other users. University teachers and other researchers can use it for their research work and analyzing the environment. High, Middle, Upper middle and middle class use it and reasonable pricing strategy will help to purchase of these notebook pc by every potential users. BEHAVIOUR FACTORS Its a general behavior of every human that NOTEBOOK PC or LAPTOP increases the status and prestige of the user. In todays busy world there is need of a product which is available with consumer anywhere or any time and at every period of time. Customer feel that they have a separate image and prestige by using ONIDA LAPTOPS. MARKET NEED FOR ONIDA LAPTOPS ONDA LAPTOPS will provides its customers the opportunity to choose NOTEBOOKs with different configuration and varieties.These laptops will also fulfill the requirements of customers and also provide benefit as:- Reasonable pricing Consumer needs a high quality product at reasonable price, for that reason ONIDA try to provide high quality product at a reasonable price. So that more and more customer can be attracted towards onida notebook pc. High quality Consumer want high quality product, which is must be high in regard to performance. ONIDA tried to meet this need of consumer at its level best. Quality was the main point to be taken care of by the company as this the main motto of company. Different varieties Choices of customer vary from person to person .So ONIDA provides different laptop with different features. And provide laptops with different configuration. Product availability One of the important things to be taken care of is Customer and customer seeking availability of product so that they can buy the product at any time they need. MARKET TRENDS ONIDA will produce different type of laptop which not only fulfill quality demand of customer but also reasonable price. So that more and more customer will purchase these laptops. Now a days NOTEBOOK PC production company is growing and more competitors are coming in the market. Now even customer are more aware about the product for that reason they want different types of product with different features and different styling features. The market for LAPTOPS with new and new technologies is growing faster, competition are increasing in these markets. Thus, ONIDA made segments for their product. They divided their customer in to three groups, and give emphasis on each of the group. So that company can capture all kind of customers in market. MARKET GROWTH Gradually total ONIDA Company is going upward with a strong competition. With the increase in domestic market, our international market growth statistics is increasing day by day. Now with the launch of new laptops with newer technology the market share of our company is going to increase further. COMPETITIVE STRATEGIES Onida will try to offer more value offerings to the customers and would try to manage CUSTOMER RELATIONSHIP. The firms in order to be effective and ahead of competition would try to maximize benefits to the consumers. Firms would react strongly to price changes made by the COMPETITORS and organizations would look at INTEGRATION on BACKEND as well as the FRONTENDS. CURRENT MARKET SITUATION OF LAPTOP MARKET Market Overview Luxury goods are now being perceived as necessities as now customers are having higher disposable incomes being spent on lifestyle products. There is a discernible shift in the consumers preference in favor of higher-end, technologically superior branded products, the demand being spurred by increasing consumer awareness and preference for new models now the modern educated customer is not confined to old technologies and old products. Now customers want to try newer products with newer technologies. Quality products with superior technology and technology up gradation have helped the industry to achieve higher growth in terms of volume and also higher realization in value terms. Rate of growth in production has been more in terms of quantity or in volume growth rather than the growth in value terms for a number of products. This has happened because of constantly falling prices over the years due to competition among the MAJOR PLAYERS, AGGRESSIVE MARKETING STRATEGIES AND DECLINING IMPORT TARIFFS. Competition has forced companies to offer efficient AFTER SALE SERVICE and support and this, in turn, has swayed customer preference for good brands. There are positive growth trends in Technological goods segments white goods and consumer electronics during and points to sustained growth because of emerging opportunities and strong fundamentals of the economy. Because of growth in production in the organized segment and domestic availability of branded products due to lowering of import duties and other liberal measures, the share of unorganized segment has come down sharply to only 8 to 10 per cent from 40 to 50 percent. The price difference between branded and unbranded goods has narrowed down and with branded players providing good after sales services and support consumer prefers to buy branded products. The industry related to technology appears to have two clearly differentiated segments. The MNCs have an edge over their Indian counterparts in terms of technology combined with a steady flow of capital. The domestic companies compete on the basis of their well-acknowledged brands, an extensive distribution network and an insight into local market conditions. Competitive strategies revolve around strong brand differentiation and prices. Bargaining power of customers is high due to availability of many brands. Demand is Cyclical and seasonal. Demand is high generally on the basis of requirement of customers as corporate customer or end user. Demand for technology is present throughout the year. There is no preference on the basis of any special month when demand for technology is more in one month than other month. Rural India which accounts for nearly 70% of the total number of households, offers plenty of scope and opportunities for the white goods industry. Increasing consumer awareness and preference for new models have added to the demand in rural areas also. And evolution of education had made more and more people rely on technology. Attractive consumer loan schemes with reduced interest rates over the years by the financial institutions and commercial banks and the hire-purchase schemes have added to the surge in demand. Besides, the consumer goods companies are themselves coming out with attractive financing schemes to consumers through their extensive dealer network. The usage of internet by the market functionaries has lead to intelligence sales of the products. It has helped sustain the demand boom witnessed recently in this sector. SWOT ANALYSIS (STRENGTHS, WEAKNESSES, OPPORTUNITIES, THREATS) STRENGTHS Quality product and services. Popular brand name. Continuous research and development Strong existing distribution channel. Increased share of organized sector as compare to unorganized sector. Attractive design. Body made of silver and plastic and available in many colors. Most of the buyers are satisfied. Presence of established distribution networks in both urban and rural areas. WEAKNESSES Poor government spending in infrastructure and other related things. Lack of promotional activities by companies. Low purchasing power of consumers Insufficient capital. Legal and political barrier regarding import, tax and shipment and other difficulties. Difficulties due to competitors. Old technologies. OPPURTUNITIES Low penetration levels Demand of NOTEBOOK PC is increasing. Improved market portfolio. Greater demand due to changing lifestyles Increasing sales throughout the country and internationally. Availability of easy finance. Promotional activities to increase brand image. Growing disposable income of consumers Increasing demand in rural sector. THREATS Strong competitors Legal political difficulties regarding import duty, tax etc. Entrance of new competitors in the market. The price of NOTEBOOK or laptops is decreasing continuously. Cheap Imports from Singapore, China and other Asian countries Higher import duties on raw materials imposed in the Budget 2008-09. COMPETITORS Onida has established its own market. But still they have to face the competitors. Some of the major competitors of onida laptops are: HP: HPs NOTEBOOK is having maximum market share in capturing the market of NOTEBOOK PC in India as well as internationally. HPS notebooks are produced in China. They are holding a leading position for long time in NOTEBOOK PC in our country. ACER: Acer is the Chinese producer of NOTEBOOK or laptops. ACER has its maximum market share of NOTEBOOK PC in Bangladesh. They have fewer shares in Indian market as compare to acre. COMPAQ: Mitsubishi is Taiwan Company. COMPAQ was an automobile company initially. Now it is producing NOTEBOOK PC also. It is now giving a strong competition to notebook or laptop market. DELL: Firstly they produce televisions. But now they have enhances its business by producing NOTEBOOK PC. And DELL is now giving a strong competition in laptop market as well as technological market. MARKET SHARE OF ONIDA PRODUCT OFFERING ONIDA will offers different type of configuration at different price to our consumer for chooses their expected product. Following are the items:- POINTS OF SUCESS Strong quality Better customer relationship management Better service Operational efficiency and integration Effective channel management Constant product change and product mix management Image of brand and product Distribution channel Retention of customer MARKETING STRATEGIES Marketing strategies are helpful in creating awareness, interest and appeal from our target market. So that more and more market can be capture. Basically marketing strategy is based on superior performance of the following areas: Different configuration. Product quality. Delight user. MISSION The main mission is to provide quality and product at a competitive price. Growth in diversity and continuous contribute to the growth is being the main market challenger. Also To benefit society at large through Innovation, Quality, Productivity, Human Development and Growth, and to generate sustained surpluses, always striving for excellence, within the framework of law, and in nothing but the truth in which we base every action 3.2 Marketing Objective Capture the market as a market leader. To attain quantitative and qualitative leadership in the technological sector. Maintaining positive and increased sales growth than the competitor. Increase the market share by market development and services. To increase product awareness and sales by persuasive promotional activities. Becoming a globally recognized and prestigious company through synergistic business investment. Differentiation through innovation and passion through empowerment. Also cost through economies of scale and world class systems and procedures that bring in delight of stakeholders. FINANCIAL OBJECTIVE The company aims to sell one lakh units of NOTEBOOKS in first year of its launching and is expecting a target turnover of Rs.100 Crores from laptop market in first year. It expects to grow at 10-15% in next five years by satisfying customer needs through its offerings. They aim to make considerable profits and achieve economies through backend and front end integration. And maintaining double digit each year. STP (SEGMENTATION, TARGETING, POSITIONING) SEGMENTATION OF MARKET The company has segmented its target market on the basis of incomes and lifestyles. People who are well educated, have lifestyle as an important element and have high income can buy laptop. TARGET MARKET The potential consumers are separated into various segments- Corporate User End User and others. The primary marketing opportunity is sell laptops to these well defined, accessible target market segments. Corporate User: Corporate users are the users who buy the laptops for their official purpose only. The MD, GM, DGM, CEO and other. Levels in offices are the main corporate users. End User: End users are the users who buy the laptops for their personal use. Teachers, Students of private University, businessman etc are mainly consider as the End users. Others: other then as mention above. POSITIONING IN MARKET Onida is trying to maintain its position as a NOTEBOOK company. The position will be achieved by providing quality product, competitive price, and according to consumers demand and by delighting consumers. There should be experienced managers to make awareness about the product to customer. Company is also promising to offer quality offerings and better services and make a satisfied consumer as its brand ambassador. The companys will advertise its product as the synonym of truth and providing it at a reasonable price. STRATEGIES ONIDAS main primary marketing strategy is to seek and firstly create customer awareness regarding the products as well as availability of product. Other marketing strategy are- Providing Total Quality Management (TQM) Customer Orientation Providing international standard products To increase the product line as well as length as per the expectations of the consumers Competitive prices of product. MARKETING MIX ONIDAS marketing mix is comprised of the following approaches to product, price, promotion, and place or customer service. PRODUCT o BRAND NAME The name of the brand is ONIDA NOTEBOOKS. o PPRDUCT CLASSIFICATION ONIDA NOTEBOOK has five types of product. These are-W125U-T3000, W3001U-T4150, W4200-T4500, W4510U, W5520U o QUALITY OF PRODUCT ONIDA is mainly popular for the maintaining of high quality of its products. Total Quality Management (TQM) is purely practiced here. o DESIGN Attractive Design, Color, configuration, Comfortable weight. o PACKING ONIDA supply the notebooks to the users with attractive packaging. It provides special cartoon with strong handling of the NOTEBOOKs to its dealers. o Size The products size vary from to one product series to other product series o Service ONIDA provides best after sales service and take feedback for its products. PROMOTION o SALES PROMOTION ONIDA provide initially shot-term incentives to encourage and purchase or sale of its products. Occasionally company has decided to give special discounted price for its products. Company also decided to give T-shirts, caps, bags etc. to lure the customers. o ADVERTISEMENT Company decided to gives advertisement for ONIDA NOTEBOOK through newspaper, billboard, popular magazine, leaflets, sponsoring on game competition, internet etc. o PERSONAL SELLING ONIDA company is also deciding to sell laptops through personal selling but till now we havent arrange any kind personal selling. PRICE o LIST PRICING Pricing of the product are being made on the basis of technology used in the product and depending upon the various series of onida. The various pricing of various varieties of laptops are: o DISCOUNT Company has decided to allow discount facilities to all the retailers and dealers. o Payment period ONIDA will sell NOTEBOOK on credit to its distributor and retailers and after the sale e period of 2 months can be given for making payment. PLACE CHANNELS ONIDA has its own distribution channel for the distribution of their products. DISTRIBUTOR ONIDA have own distribution channel for distributing their products. Location ONIDA covers district areas of our country to capture the market for their products. TRANSPORTION FACILITIES ONIDA has its own transportation facility for its distributors. Distributors are themselves responsible for taking the products to their showrooms. MARKETING RESEARCH One of the important things is research work. Research is very vital for any company to know about current market position and also to predict future needs. Information is collected through the dealer and retailer. o Question How did you hear about our ONIDA NOTEBOOK product? note the answer and this answer can be use for the promotional activities. o Customer suggestion New feedback service and customers suggestions system to gain additional information. They want to know from the customers about- à ¼ What suggestion do you want to give to company to improve our product? à ¼ Why do you need a laptop and why onida laptop only? FINANCIALS It is very important while the launching of a product to take care of the financial overview of ONIDA NOTEBOOK related to marketing activities. ONIDA address brake even analysis (BEA), sales forecast, expense forecast, and showed how this activity are link to the marketing activity. BREAK EVEN ANALYSIS Break even is a point where cost is equals to sales. The break-even analysis below shows the number of single sales, or units, that we must realize to break-even. Analysis of breakeven point is based on the cost and the sales of a company. SALES FORECAST: Sales forecast means planning or estimation of sales so that future sales can be assessed. ONIDA thinks that the sales forecast will be achieve into the five main streams; W1250U-T3000, W3001U-T4150, W4200U-T4500, W4510U and W5520U. This will steadily increase the sales. As the advertising budget allows the target market forecast, the listed of all the potential customers get divided into separate groups. The forecasted customers group divided into various categories: Corporate Users, End Users and others. EXPENSE FORECAST Every company has to make a prior forecast of expenses. It is important to mark such forecast to limit the expenses made in future. These expenses are to be budgeted at approximately 5% of total sales for 2007-08 and 6% for 2008-09. Mainly expenses are to be tracked in the major marketing categories as:-NEWSPAPER ADVERTISEMENT, PROMOTIONAL EVENTS, BILLBOARDS, PRINTED LEAFLETS, and ADVERTISEMENT IN TELEVISIONS ETC. CONTROLING FACTORS The main purpose of ONIDA NOTEBOOKs marketing plan is to serve as a guide for the organization while launching of the laptops successfully. This plan is all about implementation and changing the business and also to make it better. In this marketing plan we look at specific implementation programs, and the details that it takes to make it happen. The following are the areas which will be specially monitored to enhance the performance of company and launching of laptops. Contingency Planning Contingencies likely to arise: Price War New Technologies in the market More Competitors Companys Strategy to counter the same: Creating value preposition Continuous innovation and product development Creating distinct brand identity