Tugas Analisis Data Eksploratif Week 3

his document presents, for each STM topic (\(k = 10\)), the tweets that are **both**: - **Strongly associated with the topic**: we require a topic probability \(\gamma \ge 0.5\) for that topic. - **Highly influential in terms of engagement**: we compute an engagement score \[ \text{engagement\_score} = \text{retweet\_count} + \text{like\_count} \] using the tweet-level metadata from the social network analysis (`uniques_final_chatgpt`). For each topic, we select the **top 20 tweets** by engagement score (breaking ties by higher \(\gamma\)), and combine them into a single interactive table. This gives us a set of **influential exemplars**: tweets that not only best express each topic linguistically, but also received substantial attention in the retweet/like economy.

Implementasi PCA dan Factor Analysis pada Max Planck Weather Dataset untuk mereduksi dimensi variabel meteorologi yang saling berkorelasi. Analisis mencakup pengujian asumsi, ekstraksi komponen utama, serta identifikasi faktor laten yang menjelaskan variabilitas atmosfer secara efektif.

This document contains the materials for the lexical and stylistic research workshop organised by the Corpus Linguistics Association of Nigeria (CLAN 2025).

Camilo Porras, Juan Manuel villegas