fcm-topic-modelling
Code for the bachelor thesis of Maria Mouratidi, for the program of Cognitive Science snd Artificial Intelligence at Tilburg University. Spring Semester 2023.
Science Score: 44.0%
This score indicates how likely this project is to be science-related based on various indicators:
-
✓CITATION.cff file
Found CITATION.cff file -
✓codemeta.json file
Found codemeta.json file -
✓.zenodo.json file
Found .zenodo.json file -
○DOI references
-
○Academic publication links
-
○Academic email domains
-
○Institutional organization owner
-
○JOSS paper metadata
-
○Scientific vocabulary similarity
Low similarity (5.8%) to scientific vocabulary
Repository
Code for the bachelor thesis of Maria Mouratidi, for the program of Cognitive Science snd Artificial Intelligence at Tilburg University. Spring Semester 2023.
Basic Info
- Host: GitHub
- Owner: maria-mouratidi
- Language: Python
- Default Branch: main
- Size: 1.1 MB
Statistics
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
- Releases: 0
Metadata Files
README.md
FCM-Topic-Modelling
Code for the bachelor thesis of Maria Mouratidi, for the program of Cognitive Science snd Artificial Intelligence at Tilburg University. Spring Semester 2023.
All content presented in this work is my original creation, and I hold the rights to it. Any form of plagiarism or unauthorized use of this work is strictly prohibited. Proper citations and references have been provided for external sources used.
Abstract
Social media are an integral part of the modern world. Much of socialization, opinion exchange, and news sharing has shifted from traditional environments to social media platforms. Analyzing social media interactions can be fruitful in discovering online community dynamics or enhancing marketing techniques. For this reason, this study suggests a viable way to extract topics of interest from social networks and bring latent topic interactions to the surface. This objective is achieved with Fuzzy Cognitive Maps, which can elegantly simulate topic evolution scenarios. The use of Fuzzy Cognitive Maps is motivated by their potential to enhance network visualizations and offer more interpretability to the findings. The suggested pipeline specifically uses Latent Dirichlet Allocation to extract latent semantic topics from Twitter posts and Association Rule Mining to establish the association weights between topics. Finally, simulation scenarios are defined and topic presence dynamics are simulated with the aid of the association weights and an evolution function. The current model suggests that introducing a medium non-linearity degree to the evolution function brings the most balance between topic dynamics. Indeed, topics with manually induced presence tend to maintain a high presence while also allowing for more popular topics to dominate. Moreover, regardless of the initial activation of each topic, all topics span a confined range of activations throughout iterations. This suggests that specific presence patterns are encoded during the weight derivation process. Finally, the limitations of this study are discussed along with potential solutions.
Owner
- Name: Maria
- Login: maria-mouratidi
- Kind: user
- Repositories: 1
- Profile: https://github.com/maria-mouratidi
Citation (CITATION.cff)
# This CITATION.cff file was generated with cffinit.
# Visit https://bit.ly/cffinit to generate yours today!
cff-version: 1.2.0
title: Simulating Twitter Topic Dynamics with Fuzzy Cognitive Maps
message: >-
If you use content from this project, please cite it using the metadata
from this file.
type: Unpublished bachelor's thesis
authors:
- given-names: Maria
family-names: Mouratidi
email: mouratidi.m@gmail.com
affiliation: Tilburg University
repository-code: 'https://github.com/codingraven/FCM-Topic-Modelling.git'