Afkar, Muhammad Ihsanul (2024) Otomatisasi Chat dan Implementasi Auto Reply pada Whatsapp API Gateway untuk Membantu Penjualan UMKM. Other thesis, Institut Teknologi Sepuluh Nopember.
Text
5027201024_Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2026. Download (9MB) | Request a copy |
Abstract
WhatsApp adalah salah satu aplikasi pesan yang sangat populer, terutama di Indonesia. Oleh karena itu, banyak pelaku UMKM menggunakan WhatsApp sebagai media komunikasi dengan pelanggan mereka. Namun seiring bertambahnya pesan dan pelangggan, mengelola pesan akan semakin sulit. Penulis membuat WhatsApp API Gateway sebagai solusi untuk mengelola WhatsApp melalui antar muka web dengan fitur broadcast untuk mengirim pesan ke lebih dari satu kontak hanya dengan sekali klik, campaign untuk pesan marketing, order untuk pemesanan produk, dan auto reply berupa chatbot untuk menjawab pertanyaan-pertanyaan pelanggan. Dataset untuk chatbot didapatkan dengan mengambil riwayat pesan WhatsApp pelanggan yang tidak bersifat sensitif kemudian dikelola dan dilabeli satu per satu. Model chatbot dibuat menggunakan algoritma deep learning LSTM dengan hasil evaluasi akurasi 97,75%. Implementasi auto reply pada UMKM berhasil menjawab 19 dari 25 pesan dengan rata-rata delay balasan auto reply sekitar 1,117 detik. Broadcast dan campaign terbukti dapat mengirim chat WhatsApp ke semua penerima secara terjadwal dengan delay antara penerima sekitar 5-7 detik. Auto reply membantu penjual UMKM sehingga tidak perlu menjawab pesan pertanyaan WhatsApp pelanggan secara manual dan fitur broadcast untuk mengirim pesan secara masal dengan satu klik serta campaign untuk kebutuhan marketing dari UMKM.
=============================================================================================================================
WhatsApp is one of the most popular messaging apps, especially in Indonesia. Therefore, many UMKM players use WhatsApp as a medium of communication with their customers. However, as the number of messages and customers increases, managing messages becomes increasingly difficult. The author created a WhatsApp API Gateway as a solution to manage WhatsApp through a web interface with broadcast features for broadcasting messages to multiple contacts with just one click, campaign for marketing messages, order for product orders, and auto-reply using chatbot to answer customer questions. The dataset for the chatbot was obtained by collecting non-sensitive WhatsApp message histories from customers, which were then managed and labeled one by one. The chatbot model was created using the LSTM deep learning algorithm, with evaluation results showing an accuracy of 97,75%. Broadcasts and campaigns are proven to send WhatsApp chats to all recipients on a scheduled basis with a delay between recipients of about 5-7 seconds. Auto reply helps UMKM sellers so that they don't need to answer customer WhatsApp question messages manually and broadcast feature to send mass messages with one click and campaigns for marketing needs from UMKM.
Item Type: | Thesis (Other) |
---|---|
Uncontrolled Keywords: | Chatbot, LSTM, NLP, Whatsapp API Gateway |
Subjects: | T Technology > T Technology (General) |
Divisions: | Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Information Technology > 59201-(S1) Undergraduate Thesis |
Depositing User: | Muhammad Ihsanul Afkar |
Date Deposited: | 24 Jul 2024 04:05 |
Last Modified: | 24 Jul 2024 04:05 |
URI: | http://repository.its.ac.id/id/eprint/108717 |
Actions (login required)
View Item |