REF: https://ph02.tci-thaijo.org/index.php/asit-journal/article/view/244679
This research explored the market positions of seven pasteurized cow’s milk brands in Thailand. Data were collected from a famous social network forum in Thailand, Pantip.com. Comments from the forum were tokenized. The results of tokenization revealed 500 keywords. These keywords were categorized into eight influential factors affecint milk consumption. Brand perceptual maps by the influential factors were constructed using Multidimensional Scaling. Sementic Network and Community Detection techniques were performed on the dataset.
The entire research process are listed below.
- Web crawler & scraping
- NLP Thai Pre-train
- NLP Thai Post-train
- Create Bag-of-word
- Create cooccurrence matrix
- Robustness check
- LDA
- Create network graph
- Find community
- Validate with external data and experts