Copyright © 2025 Authors retain the copyright of this article. This article is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

@article{164465,
        author = {P. Ruchitha and Y.Akshara                   and P.Sathwik and P.S.V.Srinivasa Rao   },
        title = {ADVANCED WEB ATTACKS DETECTION WITH DEEP LEARNING},
        journal = {International Journal of Innovative Research in Technology},
        year = {},
        volume = {10},
        number = {12},
        pages = {1788-1794},
        issn = {2349-6002},
        url = {https://ijirt.org/article?manuscript=164465},
        abstract = {Web applications are a popular target for cyber attacks because they are accessible over the network and are often vulnerable. Intrusion detection systems monitor web applications and provide alerts when an attack is detected. Current implementation of intrusion detection systems usually extract features from network packets or input string features and manually select features for analysis. However, manually selecting features is time-consuming and requires in-depth security knowledge. In addition, supervised learning algorithms need a lot of legitimate records and attack request data to identify bad and bad behavior; This is often expensive and impractical for developing web applications. This article contributes to research on controlling system access. First, we evaluate the feasibility of unsupervised/semi-supervised web attack detection based on the Robust Software Modeling Tool (RSMT), which works in a supervised and behavior-oriented manner on the website. Second, we describe how RSMT trains stacked denoising autoencoders to encode and reconstruct call graphs for end-to-end deep learning; where a low-dimensional representation of the raw data of unlabeled requested data is used to identify defects using defect data. Third, we analyze the results of experimental evaluation of RSMT on synthetic data and production practices with adverse effects. Our results show that the application can effectively and accurately detect attacks including SQL injection, cross-site scripting, and deserialization with minimal information.},
        keywords = {},
        month = {},
        }