Data Mining Security And Crime Data: Measures And Role Of Clustering Method

Data Mining on the Crime Domain

Discuss about the Data Mining Security and Crime Data.

Data mining is referred to as an advanced technology that is widely used in different business organizations to extract hidden but predictive information from the data server. In order to make knowledge driven decisions data mining tools are needed to be used appropriately by the organizations. While implementing the data mining tools in different business organizations, different security measures are required to be considered by the researchers as the data stored in the data server are all very much sensitive. In case, data is hijacked by external users then that is considered as crime. In order to conduct the research the background of the problems, aim, objectives, a previous study, research methodologies and the collected data will be analyzed by the researchers. 

In order to aid the procedure of crime as well as criminal investigation, different crime control applications are developed. Certain extracted factors such as seriousness, frequency, duration are used to make comparison between the criminals by measuring the data accordingly. According to Ismail et al. (2013), a regional crime analysis program should be proposed to mitigate the challenges. While extracting data from the server, proper security measures are needed to be adapted to be adapted by the users as well as by the owners (Sood, Garg and Palta 2016). Particularly for crime data analysis, the data mining approach is undertaken as an algorithm. Depending on the profiles, that should be analyzed to determine the criminals and their crime, proper comparison is also conducted. For accelerating the crime solving process, many developers have used K-mean clustering crime pattern detector (Win, Tianfield and Mair 2014). Two-phase clustering algorithm or AK mode is also used by some developers to find out automatically similar case subsets from large dataset (Shea and Liu 2013).

In order to produce high quality data mining results, data processing technique is found to be very much useful. As, raw data are collected from different resources and get stored into database management system or data warehouse, thus those data are needed to be well processed for further important usage (Wang et al. 2015). The steps of data processing include data cleaning, data integration, data transformation and data reduction. Sagiroglu and Sinanc (2013)           stated that in the initial phase, the missing data are collected and the noisy data are smoothened, unnecessary data is removed and the conflicting data are resolved. In the integration phase, the information are integrated together with normalization and transformation, to get better results regarding the data. 

Role Played by Data Processing Technique

Three main classification of partition clustering method are k-means, AK-means and the expectation maximization method. Depending on the mean values, the data clusters are segmented with the help of K-means. From the internal distance among the objects, the mean value is calculated (Ward and Peppard 2016). The steps of k-mean and AK-mean algorithms are completely separate from one another. Sood, Garg and Palta (2016) stated that, in order to find out the mean value, K-means algorithm acts as a base for the clustering algorithm; whereas, for automatically finding same subsets from a large data set, AK mode algorithm is widely used.  

For serving the crime investigation appropriately, data mining provides active solutions to the users and service providers also. In order to detect crime data mining technology is used by different business organizations (Teller, Kock. and Gemunden 2014). For automatically associate objects into the crime records concept space clustering technique is used by the developers. In order to detect fraud different abnormal activities deviation detection technology is used. (Jaber et al. 2015). For detecting email spamming, classification approach and to perceive deceptive data in the criminal records string comparator has been widely used. Moreover, it can be said that security in data mining approach and detection of crime data are necessary to understand the challenges very easily.

According to Ismail et al. (2013), security is referred to as one of the major concern that is required to be considered by different business organizations, those are using data warehouse or normal database management system to store the confidential information. For the current business perspective, data theft is another major issue. In case of data mining approach, the security of data server is one of the major concerns. While extracting data from the storage areas with the help of data mining tools, proper authentication is required to be considered so that the unauthenticated users are failed to hijack private information. Once the data get theft is called cybercrime, because the hackers hacked the sensitive information from the storage area.    

Figure 1: Data processing

(Source: Wang et al. 2015, pp. 310)

H_0: Security in data mining is needed be incorporated for the consumer security.

H_1: Security in data mining is not needed be incorporated for the consumer security.

Before the initiation of research, it is very much important to select suitable methodology for data mining security and crime data. The selection methodology deals with research philosophy, approach, research design and data collection method.

Clustering Method for Crime Domain

Based on the category of the topic – “data mining security and crime data” the researcher should adapt positivism research philosophy and not interpretivism. This particular philosophy helps to accomplish the research work by developing hypothesis, after considering previous studies. For this case, the researcher needs not to develop new frameworks and theories. This research work is focused on the need of security in data mining to reduce the range of cybercrime all over the world.                           

Three different types of research design approaches such as descriptive, explanatory and exploratory are used by researcher while developing academic research works. In this case, the researcher must consider descriptive research design and not explanatory and exploratory as it deals with previously mentioned research aim and issues very clearly.

Two different types of research approaches such as deductive and inductive are used by the researchers while developing research proposal. Inductive approach is time taken and is conducted by using secondary data. In this case, the researcher needs to complete the research work within a specified time thus; the researcher needs to adapt deductive research approach.

For different research works, two different kinds of data collection methods such as primary and secondary are found. The data those are collected from the similar field of study are referred to as primary data whereas; if the data are collected from various journals, Articles, books then that is known as secondary data collection approach.

Based on the types and collection approaches, the data are divided into two different forms such as qualitative and quantitative data. Statistically analyzed data are quantitative data and the data those cannot be analyzed statistically are called qualitative data (Treiman 2014). For this research topic, both qualitative and quantitative data collection approaches will be used by the researcher. Quantitative data will be collected from different survey and qualitative data will be collected by conducting direct interviews with top-level management authorities of different business organizations that uses data mining technique to extract information from the data server.

Sampling is referred to as a procedure where, a particular part from the entire population is selected under the study area of the research topic. Probability sampling and non-probability sampling these two types of sampling technologies are used by different researchers. For this case, Probability sampling technique will be used by the researcher.

Before conducting the research, the researcher must consider appropriate time schedule that is provided in the following table. However, with the changing requirement some of the schedules might be changed during the progress period.

The Research activities	Week 1	Week 2	Week 3	Week 4	Week 5	Week 6
Proper understanding of the problem area of data mining security and crime data	ü
Literature review		ü	ü
Research methodology			ü	ü
Data collection				ü	ü
Sampling and Data analysis					ü
Research findings					ü
Conclusion						ü

Data Mining as an Active Investigator for Crime Solving

Conclusion     

 From the overall discussion, it can be concluded that in order to identify the crime pattern. The usage of data mining technology is increasing at a rapid scale. Data mining tool is used to detect crime data but cannot be used to replace the risks from the core business area. Currently most of the medium to large organization uses the concept of data warehousing to store data with security. Technologies provide high range security measures but still due to certain lags in the organizational goal, the data stored in the database or data warehouse might be hacked and misused by the criminals. In order to mitigate these security relevant issues, the businesses should incorporate proper encryption keys, authentication approach at the same time. The encryption key will encrypt the data and thus only the authenticated users will be able to get the access of the data and none of the unauthenticated users will be able to access it.

References

Brown, J. and Stowers, E. (2013). Use of Data in Collections Work: An Exploratory Survey. Collection Management, 38(2), pp.143-162.

Eaton, S. 2013. The Oxford handbook of empirical legal research. International Journal of Social Research Methodology, 16(6), pp.548-550.

Ismail, M.N., Aborujilah, A., Musa, S. and Shahzad, A., 2013, January. Detecting flooding based DoS attack in cloud computing environment using covariance matrix approach. In Proceedings of the 7th International Conference on Ubiquitous Information Management and Communication (p. 36). ACM.

Jaber, A.N., Mohamad Fadli, Z., Ahmed, D., Ahmed, F.D. and Mazlina, A.M., 2015. Security Everywhere Cloud: An Intensive Review of DoS and DDoS Attacks in Cloud Computing. Journal of Advanced & Applied Sciences (JAAS), 3(5), pp.152-158.

Mackey, A., & Gass, S. M. (2015). Second language research: Methodology and design. Routledge.

Masala, E., Servetti, A., Basso, S. and De Martin, J.C., 2014. Challenges and issues on collecting and analyzing large volumes of network data measurements. In New Trends in Databases and Information Systems (pp. 203-212). Springer International Publishing.

Pek, G., Lanzi, A., Srivastava, A., Balzarotti, D., Francillon, A. and Neumann, C., 2014, June. On the feasibility of software attacks on commodity virtual machine monitors via direct device assignment. In Proceedings of the 9th ACM symposium on Information, computer and communications security (pp. 305-316). ACM.

Sagiroglu, S. and Sinanc, D., 2013, May. Big data: A review. In Collaboration Technologies and Systems (CTS), 2013 International Conference on (pp. 42-47). IEEE.

Shea, R. and Liu, J., 2013. Performance of virtual machines under networked denial of service attacks: Experiments and analysis. IEEE systems journal, 7(2), pp.335-345.

Sood, R., Garg, S. and Palta, P., 2016. A Novel Approach to Data Filtration against Packet Flooded Attacks in Cloud Service. Journal of Network Communications and Emerging Technologies (JNCET) www. jncet. org, 6(5).

Teller, J., Kock, A. and Gemünden, H.G., 2014. Risk management in project portfolios is more than managing project risks: A contingency perspective on risk management. Project Management Journal, 45(4), pp.67-80.

Treiman, D. J. (2014). Quantitative data analysis: Doing social research to test ideas. John Wiley & Sons.

Wang, B., Zheng, Y., Lou, W. and Hou, Y.T., 2015. DDoS attack protection in the era of cloud computing and software-defined networking. Computer Networks, 81, pp.308-319.

Ward, J. and Peppard, J., 2016. The Strategic Management of Information Systems: Building a Digital Strategy. John Wiley & Sons.

Win, T.Y., Tianfield, H. and Mair, Q., 2014, December. Virtualization security combining mandatory access control and virtual machine introspection. In Utility and Cloud Computing (UCC), 2014 IEEE/ACM 7th International Conference on (pp. 1004-1009). IEEE.