Analiza grafa društvene mreže primjenom tehnika strojnog učenja

master's thesis

• Title (english): Social Network Graph Analysis Using Machine Learning Techniques
• Author: Ivan Despot
• Mentor: Goran Delač (mentor)
• Committee member: Goran Delač (predsjednik povjerenstva)
• Committee member: Marin Šilić (član povjerenstva)
• Committee member: Klemo Vladimir (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2021-07-06, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Ovaj rad sadrži kratak pregled područja analize društvenih mreža te znanosti o mrežama (engl. network science). Poseban naglasak stavljen je na tzv. algoritme ugrađivanja (engl. embedding algorithms) te njihovu primjenu kod standardnih tehnika analize mreža. Aplikacija, u programskom jeziku Python, implementirana je i dokumentirana za potrebe evaluacije algoritama ugrađivanja nad grafovima kao što su node2vec te DeepWalk. Evaluacija je izvedena pomoću zadataka klasifikacije čvorova (engl. node classification), predviđanja veza (engl. link predication) te grupiranja čvorova (engl. node clustering) za potrebe detekcije zajednica. Evaluacija je provedena nad javno dostupnim podacima u obliku grafova društvenih mreža te se može reproducirati nad drugim skupovima podataka u svrhe određivanja optimalnog algoritma ugrađivanja te parametara za specifične slučajeve.

Abstract (english)

This thesis provides a brief introduction to the field of social network analysis as well as network science in general. A special focus was given to embedding algorithms and their application in standard network analysis techniques. A Python application has been implemented and documented for the purpose of evaluating the graph embedding techniques node2vec and DeepWalk on machine learning tasks such as node classification and link prediction as well as community detection (node clustering). The evaluation was undertaken on publicly available social network graphs and can be reproduced on other datasets to determine the best algorithm implementations and parameters on a case-to-case basis.

• Keywords: analiza grafa društvene mreže; znanost o mrežama; algoritmi ugrađivanja grafova; node2vec; DeepWalk; strojno učenje; klasifikacija čvorova; predviđanje veza; grupiranje; otkrivanje zajednica
• Keywords (english): social network graph analysis; network science; graph embedding algorithms; node2vec; DeepWalk; machine learning; node classification; link prediction; clustering; community detection
• Language: english
• URN:NBN: urn:nbn:hr:168:185388
• Study programme: Title: Computing; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka računarstva (magistar/magistra inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2023-01-13 09:55:04