ITC516 | Security in Data Mining | Inforamtion Technology

Home ITC516 | Security in Data Mining | Inforamtion Technology

ITC516 | Security in Data Mining | Inforamtion Technology

Data mining is known as the process of posing a sequence of adequate queries to retrieve information from wide quantities of data in a database. Data mining techniques are applied to manage issues in the database security. Data mining involves those grounded in inductive logic programming, rough sets, neural networks and machine learning among others. Data mining techniques evolve applications in intrusion detection as well as auditing databases. Moreover, one can apply tools of data mining for detecting some abnormal patterns.

What are the issues associated with data mining?

Data mining is not a simple task because algorithms involve highly complex, and data is not available in one place. It requires being integrated from different heterogeneous data sources. Major issues associated with data mining are:

Mining methodology and user interaction issues

Mining various forms of knowledge in databases:

Various users can be interested in several forms of knowledge. Hence, it is essential for data mining to cover a wide range of knowledge discovery activities.

Interactive mining of knowledge at different levels of abstraction:

Data mining process requires being interactive as it enables the user to emphasize the search for patterns, providing and improving data mining requests by returned results.

Data mining query languages as well as ad-hoc mining:

Data mining query languages which enable the user to demonstrate ad-hoc mining tasks must be integrated with the data warehouse query language.

Managing noisy or incomplete data

The techniques of data mining are desired to manage noise as well as incomplete objects when mining the regularities of data. When data cleaning techniques are not present, then the accuracy of determined patterns will be poor.

Performance issues

Scalability and efficiency of data mining algorithms:

To efficiently retrieve the information from wide range of data in the databases, data mining algorithm should be scalable and efficient.

Distributed, parallel and incremental mining algorithm:

The factors like wide database size, a large distribution of data and difficulty of data mining techniques motivate the development of distributed and parallel data mining algorithms. These algorithms divide the data into segments that are further processed in the parallel fashion.

Diverse data types issues

Management of relational and complex data types:

The database may involve complex data objects, spatial data, multi-media data objects and temporal data. It is not probable for one system to mine all forms of data.

Mining information from the heterogeneous database as well as global information systems:

Data is accessible from various data sources on WAN or LAN. This data source might be structured, unstructured or semi-structured. Hence, mining knowledge from them adds issues to data mining.

Best practices for securing data mining

Granular access control: Access control is related to two aspects as per CSA, i.e. restrict user access and grant the user access. To setup, granular access control defines quick hit tips defined beneath:

  • Maintain access labels
  • Trace secrecy requirements and assures proper implementation
  • Involve single sign-on

Audit: Granular auditing is must in data mining security specifically after the attack on a system. CSA recommends enterprises to generate a cohesive audit view. Audit information confidentiality and integrity are also necessary.

Endpoint filtering and validation: End point protection is paramount for an enterprise by started using trusted certificates and performs resource testing. Statistical similarity detection techniques can be involved to filter malicious inputs while protecting against Sybil attacks.

Secure data storage and transaction logs: Storage management is a crucial part of the security equation. CSA recommend using signed message digest for providing a digital identifier for digital documents and involve a technique defined secure un-trusted data repository for detecting unauthorised file modification performed by malicious server agents.

Solve your assignments with the help of online professionals

One of the notable and most essential features of an assignment is authenticity factor of assignment. The masters ensure that assignment delivered is free from grammatical errors and plagiarism. Receiving quality at an affordable rate is a dream come true. Solving assignments with the help of online professionals enables scholars to satisfy the punctuality criteria and helps in submitting assignments within a stipulated time.

Latest Reviews

  • Fletcher
    27 Jan, 2019

    A big thanks to the writers who helped me in my business essay. My essay was completed in the specified time period. All my lecturers were very satisfied with my work.

  • Jasper
    12 Jan, 2019

    This is the best site ever. It provided with the best help for my computer assignment. My result was positively affected because of effective assignment work.

  • Sofia Saleem
    31 Oct, 2018

    I must say that I am 100% satisfied and impressed with the hard work and efforts that your writers have put in. I would love to recommend your services to my friends and class mates. Great work!

View All Reviews