PERANCANGAN MODEL EKSTRAKSI TOPIK DATA ULASAN KONSUMEN PT X MENGGUNAKAN METODE TEXT MINING
PT X is an online travel agent (OTA) company. PT X’s total transacting customer declined on February 2022 as well as April 2022. The decline was caused by the increase of customer dissatifaction that is reflected on ratio increase in review with lower ratings. The root cause of this problem is produ...
Saved in:
| Main Author: | |
|---|---|
| Format: | Final Project |
| Language: | Indonesia |
| Online Access: | https://digilib.itb.ac.id/gdl/view/69088 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Institut Teknologi Bandung |
| Language: | Indonesia |
| Summary: | PT X is an online travel agent (OTA) company. PT X’s total transacting customer declined on February 2022 as well as April 2022. The decline was caused by the increase of customer dissatifaction that is reflected on ratio increase in review with lower ratings. The root cause of this problem is product development that isn’t in accordance to customer voice led by absence of tools to extract customer voice from customer review.
This research is conducted to build topic modeling for PT X’s customer review. The Cross-Industry Standard Process Method for Data Mining (CRISP-DM) is framework used to build the model. Data that is used for this research is 27.085 rows PT X’s of customer review from Google Play Store. Models that are built for this research are Latent Dirichlet Allocation (LDA) and Latent Semantic Analysis (LSA).
From the conducted research, LDA with 8 topics is the best model because it has the highest cohrence value which is 0.478. Moreover, this research also generated model visualisation using pyLDAvis along with topic prevalence score for each topic on model deployment. Output from this research is expected to be used as an input for PT X’s product development based on topics from customer review.
|
|---|