Penggunaan Metode Term Weighting untuk Filtering Dalam Object Based Opinion Mining pada Review Produk Pariwisata

Afrizal, Ahimsa Denhas (2019) Penggunaan Metode Term Weighting untuk Filtering Dalam Object Based Opinion Mining pada Review Produk Pariwisata. Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 05211550010015-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2023. Download (1MB) | Request a copy

Abstract

Review memiliki peran penting dalam industri pariwisata, terutama industri hotel dan restoran. Opinion mining diperlukan untuk mengolah data review yang sangat banyak, sehingga menjadi lebih berguna bagi pelaku industri pariwisata.
Penggunaan metode filtering terbukti dapat meningkatkan hasil ekstraksi fitur dari object based opinion mining, dan secara umum dapat meningkatkan hasil klasifikasi opini dari review produk pariwisata. Namun, belum ada suatu metode penyusunan data filter yang baku, sehingga penyusunan data filter selama ini hanya ditentukan oleh ahli.
Penelitian ini mencoba menerapkan beberapa metode term weighting yang berbeda, seperti TF-IDF, mTFIDF, dan BM25 untuk menyusun data filter secara otomatis. Data filter yang disusun secara otomatis ini akan dinilai oleh ahli, dan akan dibandingkan hasilnya dengan data filter yang disusun oleh ahli. Hasil dari penelitian ini adalah metode term weighting terbaik yang dapat digunakan untuk menyusun data filter secara otomatis, sehingga akan meningkatkan performa dari proses object based opinion mining.
Hasil dari object based opinion mining dengan data filter term weighting, menunjukkan bahwa TFIDF menjadi metode term weighting terbaik apabila dikombinasikan dengan frequent object dengan accuracy 37.98%, precision 50.69%, recall 44,28%, dan fmeasure 47.27% untuk data hotel. Sedangkan untuk data restoran memberikan hasil accuracy 37.98%, precision 50.69%,recall 44.28%, dan fmeasure 47.27% untuk data restoran.
================================================================================================
Reviews is an important thing in tourisms, especially for hotel and restaurant industry. Opinion mining used for processing a huge amount of review data, so it can be more useful for the industry.
The utilization of filtering can improve the feature extraction result from object based opinion mining, and can improve opinion classification result generally. However, there is no proven method yet to develop filter data automatically. All the time, the method for developing data filter, only depends on judgement by an expert.
This research try to apply several term weighting method such as TF-IDF mTFIDF and BM25 to develop filter data automatically. The filter data result from term weighting, will be compared to the filter data developed by the expert. The result from this research is the best term weighting method for developing filter data, that can improve the feature extraction and opinion mining relatively.
TFIDF become the best term weighting method applied for filter data combined with frequent object, the result are accuracy 37.98%, precision 50.69%, recall 44,28%, and fmeasure 47.27% for hotel data. Meanwhile for restaurant data, the result are accuracy 37.98%, precision 50.69%,recall 44.28%, and fmeasure 47.27%.

Item Type:	Thesis (Masters)
Additional Information:	RTSI 006.312 Afr p-1 2019
Uncontrolled Keywords:	Opinion mining, feature extraction, filtering, term-weighting.
Subjects:	Q Science > QA Mathematics > QA76.9.D343 Data mining. Querying (Computer science) T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7872.F5 Filters (Electric) Z Bibliography. Library Science. Information Resources > ZA Information resources > Z699.5 Information storage and retrieval systems
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Information System > 59101-(S2) Master Thesis
Depositing User:	Afrizal Ahimsa Denhas
Date Deposited:	06 Oct 2021 19:19
Last Modified:	06 Oct 2021 19:19
URI:	http://repository.its.ac.id/id/eprint/61326

Actions (login required)

View Item