Tutorial Code Repository for Data Science for Communication & Social Networks at USC

Class by: Luca Luceri
Current TA: Eun Cheol Choi

Former TAs/contributers:

• Meiqing Zhang
• Herbert Chang
• Alex Bisberg
• Emily Chen
• Julie Jiang

Past repositories: https://github.com/herbertfreeze/COMM557, https://github.com/echen102/COMM599

Fall 2024 Setup

• Google CoLab: To circumvent dependency issues, all in-class tutorial scripts will be tested on Google CoLab.
• Anaconda/Jupyter: Optional, but highly recommended.

CoLab links to tutorial scripts:

Tutorial 0 Python Crash Course

• https://colab.research.google.com/drive/172g1iuYfTbKFhKuhPJHt1us1EAPJazUx?usp=sharing
• https://colab.research.google.com/drive/1cbrKLErdEKpiCIvb9vDe2Q2gyKoeH6q5?usp=sharing

Tutorial 1 Analyzing and Visualizing Networks with 'networkx'

• NetworkX: https://colab.research.google.com/drive/1dGWSYjfMW_nA3DDIi7OS0iWLGMiCrghb?usp=sharing

Tutorial Gephi

• Les Miserables file derived from https://github.com/gephi/gephi/wiki/Datasets
• Airlines-sample file derived from https://github.com/gephi/gephi.github.io/tree/master/datasets

Resource:

• Gephi Tutorials: https://gephi.org/users/
• Gephi Datasets: https://github.com/gephi/gephi/wiki/Datasets
• Gephi Wiki: https://github.com/gephi/gephi/wiki/

Tutorial 2 Reddit Scraper & Analyzing Reddit Networks

• Reddit Scraper: https://colab.research.google.com/drive/15g4wbv4-J_fIUbI8TZ3QqNYBokMHzD0p?usp=sharing
• Analyze networks: https://colab.research.google.com/drive/1LpA_tMeJpusUmbIuESwSQ_l3awEqFmjj?usp=sharing

Tutorial 3 Analyzing Wikipedia Networks

• https://colab.research.google.com/drive/19HZGu5f7cOmxDekOxaa2Eax79SVjZU-n#scrollTo=fICebxnSiKjr

Tutorial 4 NLP (1): Preprocessing, Sentiment Analysis, etc.

• Basic preprocessing, Named Entity Recognition, Word cloud, Semantic network: https://colab.research.google.com/drive/1mXI95F_2RGnZjzWyVGp5cK3c3b87MtZI?usp=sharing
• Sentiment analysis using Naive Bayes Classifier, VADER: https://colab.research.google.com/drive/1lHX9l0Z4FiWw0HdIl6elTDcWyNVCL6HM?usp=sharing
• Sentiment analysis using RoBERTa: https://colab.research.google.com/drive/17jqpyh4ZW2OmQsO_giZi7l0QlQWu_mgR?usp=sharing

Tutorial 5 NLP (2): Topic Modeling

• LDA visualization: https://colab.research.google.com/drive/1_BfXYlVwZVmCdiGD6Sb2ZW31vBK7-RDI?usp=sharing
• BERTopic: https://colab.research.google.com/drive/1STD-3oajZQguZj4Zvu5yhHVQlEEWdXtd?usp=sharing

Tutorial 6 LLMs, GPTs

• Various NLP tasks with BERT-based LMs: https://colab.research.google.com/drive/1wIKIhkNy-y-QZtF6JYzowZT4BeOqlSDp?usp=sharing
• LLMs with model weights publicly available: https://colab.research.google.com/drive/1C0SqKDV9QRO8DIE9_VY6OkfzJbFJT2qL
• OpenAI API: https://colab.research.google.com/drive/1wPrJuG_J36fA-K4PqIJip4xrBpWIPe0O

Tutorial 7 Supervised learning

• Iris (classification task) & Housing (regression task): https://colab.research.google.com/drive/16nSXJ4U6XoRD6oQoZeyYYlWocaU9toTi?usp=sharing
• Misinformation classification using tree-based models: https://colab.research.google.com/drive/1J-KXtEvo_G__QfusVoKQUSPGouTsERlr?usp=sharing
• Feature selection & extraction: https://colab.research.google.com/drive/1O1BINMabdN2V8wksPZsixkHg14Vs49ZJ?usp=sharing

Tutorial 8 Supervised learning by fine-tuning BERT-based models

• Binary task: https://colab.research.google.com/drive/19vVMycNAEEwLeF25jumTUY3qVaCpYx5C?usp=sharing
• Multiclass: https://colab.research.google.com/drive/1Ulu0m_H_VA8wX2SswK26aLTsMmPI7FZ2?usp=sharing
• Cross-validation: https://colab.research.google.com/drive/1j8ORY_0qB9bGzOMFrAMUEOe1Y9dHnCBh?usp=sharing

Tutorial 9 Unsupervised learning

• Clustering (k-means, HCA, DBSCAN, GMM) and dimensionality reduction (PCA, t-SNE, AE) using misinformation dataset: https://colab.research.google.com/drive/1byLOGWYJrMEvHc6joGm66M2tjnYxxN6r?usp=sharing

Additional resource

• OpenAI Batch API: https://colab.research.google.com/drive/1A8tj4g3pZFjxY22368U_VGylcC6QjbDx?usp=sharing