<![CDATA[Lhokseumawe a city known for its numerous coffee shops, serves as the focus of this study, which aims to develop a coffee shop recommendation system using a content-based filtering approach based on Google Maps review analysis. A total of 54 coffee shops were collected through web scraping and filtered to 32, as only these shops provided sufficient and relevant reviews according to the selected keywords. User reviews were processed through preprocessing, TF-IDF weighting, and cosine similarity to measure the alignment between user preferences and shop characteristics. A scenario-based evaluation was conducted by using keywords such as “noodles,” “parking,” “spacious,” “toilet,” and “watching together” to represent user preferences. The results show that the system generates recommendations consistent with the presence and relevance of these keywords, with shops such as AN Coffee and Arabica Kopi frequently appearing as top suggestions. Although the evaluation is limited to scenario-based testing, the system demonstrates potential in assisting users in selecting suitable coffee shops. Future work may include hybrid filtering, machine learning methods, automated keyword extraction through topic modeling, and user-based evaluation to improve recommendation quality.]]>
Copyrights © 2026
