KOMPARASI ALGORITMA KLASIFIKASI DAN EKTRAKSI FITUR PADA SISTEM AKUNTANSI PINTAR

Bagas Adi Makayasa, NIM.: 21206052011 (2023) KOMPARASI ALGORITMA KLASIFIKASI DAN EKTRAKSI FITUR PADA SISTEM AKUNTANSI PINTAR. Masters thesis, UIN SUNAN KALIJAGA YOGYAKARTA.

Preview	Text (KOMPARASI ALGORITMA KLASIFIKASI DAN EKTRAKSI FITUR PADA SISTEM AKUNTANSI PINTAR) 21206052011_BAB-I_IV-atau-V_DAFTAR-PUSTAKA.pdf - Published Version Download (3MB) | Preview
	Text (KOMPARASI ALGORITMA KLASIFIKASI DAN EKTRAKSI FITUR PADA SISTEM AKUNTANSI PINTAR) 21206052011_BAB-II_sampai_SEBELUM-BAB-TERAKHIR.pdf - Published Version Restricted to Registered users only Download (4MB) | Request a copy

Abstract

The problem of financial records that are not in accordance with the principles of accounting science has the potential to cause unnecessary problems. Micro, Small and Medium Enterprises with their characteristics still encounter many obstacles in writing financial reports. This research aims to explore opportunities for implementing an accounting automation system using the NLP approach, interpreting financial transactions based on the text written on the transaction form into accounting journals (debits and credits). Experiments were carried out by comparing the performance of three classification algorithms (namely SVM, K-Nearest Neighbor, and Random Forest) with Feature Extraction (TF-IDF, BOW, and Word2Vec). There are 200 financial transaction datasets consisting of ten classes, the data is divided into two parts, namely balance dataset and imbalance dataset. The SVM and Word2Vec pair in the balance dataset gave the highest score accuration (92.5%), precision (92.5%), recall (93.33%), and F1 score (92%). However, when compared with the results of related semantic research (the average performance reaches 95%), the results obtained in this study are still lower. One thing that may have a significant effect is the amount of data in the corpus which is still lacking. Researchers suggest increasing the number of datasets and trying to use a combination of language models such as Glove, Bert etc. This research can also be used as an initial model for more complex financial transaction cases in future studies.

Share this knowledge with your friends :

Actions (login required)

View Item