Institutional Repository of Vilnius University
A system for detecting cryptocurrency token scams /
Abstract
dc:descriptionThe aim of this thesis is to design and implement "Tokenomics Lab," a holistic web-based information system dedicated to automating the risk analysis of cryptocurrency tokens across multiple blockchain networks. The research addresses the problem of fragmentation in the DeFi sector, where retail investors lack unified tools to detect scams like "rug pulls" and "honeypots." To achieve this aim, the following objectives were set and accomplished: investigating common tokenomic risk factors, designing a custom weighted 10-factor risk scoring algorithm, developing the application using a serverless architecture, and validating the system's accuracy. The research methodology followed an Agile/Scrum development framework. The system was built using a modern technology stack, including Next.js 16 for the frontend, Firebase for the backend database and authentication, and the Groq AI SDK (Llama 3.3) for natural language processing. A "Chain-Adaptive" architecture was implemented to normalize data from disparate sources (Ethereum, Solana, etc.). The system’s performance was evaluated through algorithmic simulations against a dataset of known malicious and safe tokens, as well as qualitative beta testing with users. The main result is a fully functional platform that aggregates real-time market and security data to generate a comprehensive risk profile. The system successfully calculates a risk score on a 0–100 scale and provides AI-generated narrative explanations. Testing demonstrated that the risk algorithm achieved an 80% accuracy rate in identifying scam tokens, validating the effectiveness of combining algorithmic scoring with Large Language Models. The "Tokenomics Lab" platform successfully bridges the information gap in the DeFi market. It proves that complex on-chain data can be processed and simplified for retail investors without sacrificing analytical depth, providing a viable solution for automated fraud detection. Scope of the thesis: The thesis consists of [104] pages, [16] tables, and [30] figures. Keywords: Cryptocurrency, Tokenomics, Fraud Detection, DeFi, Artificial Intelligence, Risk Analysis.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2026
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Maniyath Joshy, Nayan Joshy,
Subjects
dc:subject × 1Rights
dc:rights- Statement dc:rights
-
- info:eu-repo/semantics/openAccess
- Language dc:language
- eng
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.vu.lt/VU:ELABAETD277690681&prefLang=en_US