| Abstrakt: |
In Modern era the Internet has play a vital role and called as modern revolution in terms of the speed and scales of both storage and computation means thereby opening a gateway to many further innovations and research domains. This increasing boundary of computation also brings along an alarming concern of Cyber Security along with it. Cyber Attacks intended at infrastructures, networks, personal computer devices, targeting information available in the systems and using various tools to change, remove or steal data or information systems are coming into light. One such prominently attempted attack is the SQL Injection Attack. Injection Attack by Structured Query Language (SQL) (IASQL) maintains an attacker's preferred method of stealing confidential data from a database, with potentially disastrous effects, on insecure web applications. Further IASQL can effect in deals with non authorized to use secure data, such as credentials, banking information details, or user information. Most high priorities data has violated in recent days because of query based injection attacks, and it has significant loss and rigid of its action. An attacker can sometimes get a permanent gateway into an organization system, resulting in a long-term penetration that goes unreported for a long time. One approach that would be suggestible to be followed to mitigate SQL Injection attacks is the use of classification to bias safe queries from tweaked queries to detect SQLI attacks. However, IASQL research, there is a lack of predefined, reliable data or collection of data with features and past data items to train a classifier and a majority of detections happen observation in such cases. Through this study, we intend to deploy a web application that makes use of template formats of Injection attacks by SQL by modeling the query symbols and token which has been available to perceive the affected area if a given string or a parameter of a query would end up tampering with the query itself. This would essentially mean that we would define an intelligent system that could detect an anomaly in the query that may make it a threat on production. In this prototype model, we built an application using the technique of Classification and tokenization to propose a new system to detect spiteful websites over the internet. [ABSTRACT FROM AUTHOR] |
