Scribd
Upload
125 views

Dark Web Classification Based On Text CNN and Topic Modelling Weight

Major project 3

Uploaded by

nithish14556
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
125 views

Dark Web Classification Based On Text CNN and Topic Modelling Weight

Major project 3

Uploaded by

nithish14556
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as DOCX, PDF, TXT or read online on Scribd
You are on page 1/ 3

Dark side of the web : Dark web classification based on text CNN and

Topic modelling weight

Abstract
The "dark web" represents a hidden portion of the internet, accessible only
through specialized software and often associated with illegal or
unregulated activities. This project explores a method for classifying dark
web content using a combination of Convolutional Neural Networks
(CNN) and topic modeling techniques, providing insights into text-based
content across dark web sites. The proposed approach utilizes topic
modeling to extract thematic representations of textual data, creating
weighted topic vectors that feed into a text CNN model. The CNN-based
classifier, trained on these enriched features, achieves robust and accurate
classification of dark web content into categories such as marketplaces,
forums, or illicit services. By enhancing visibility into dark web activities,
this model supports law enforcement, cybersecurity, and research
communities in understanding and managing dark web risks.
Introduction
The dark web hosts vast, anonymous, and often illegal content that poses
significant challenges for law enforcement and cybersecurity experts.
Traditional approaches to content analysis struggle due to the unstructured
nature and anonymity of data sources. This project introduces a CNN-
based classification model augmented with topic modeling weights to
improve the identification and categorization of content, helping
distinguish between various types of dark web activities.
Existing System with Disadvantages
Conventional methods of dark web content analysis, such as keyword
searches and manual data categorization, are time-consuming and limited
in scope, often failing to capture the full thematic context of dark web
texts. While some machine learning techniques are used, many lack the
capacity to handle the unstructured and high-dimensional nature of dark
web data, reducing classification effectiveness and accuracy.
Proposed System with Advantages
The proposed system leverages a hybrid approach combining topic
modeling and CNN. Topic modeling captures thematic information from
text data, creating a context-rich representation. This is then weighted and
used as an input to a CNN model that excels at extracting complex
patterns. By integrating thematic weights into the CNN classification
model, this system enhances classification accuracy, scalability, and
robustness in detecting diverse types of dark web content, thus enabling a
proactive approach to dark web monitoring.
HARDWARE & SOFTWARE REQUIREMENTS:

HARD REQUIRMENTS :

 System : i3 or above.
 Ram : 4 GB.
 Hard Disk : 40 GB

SOFTWARE REQUIRMENTS :

 Operating system : Windows8 or Above.


 Coding Language : python
 Idle : Jupyter Notebook

Conclusion
This project presents an innovative classification approach for dark web
content using CNN and topic modeling. By effectively categorizing dark
web activities and creating actionable insights, the model supports
cybersecurity and law enforcement efforts to monitor, understand, and
mitigate the risks associated with the dark web. This approach holds the
potential to significantly improve the visibility of dark web activities,
helping address the challenges posed by this hidden segment of the
internet.

You might also like