Dataset of Understanding Guest Review From Google Play Using Naïve Bayes-Based Data Analysis: A Study on Nanovest

This dataset contains user reviews of Nanovest, an investment application for AS stocks, gold, and cryptocurrency. The data was collected from user reviews of the Nanovest app on Google Play. The reviews, written in Indonesian, reflect users' experiences and opinions regarding the app’s features, security, and functionality. By analyzing this review data, this study aims to determine the proportion of positive and negative reviews and identify the key aspects frequently mentioned by users. The findings of this study can provide recommendations for improving service quality and application performance for cryptocurrency investment platforms in Indonesia.

This dataset was collected through web scraping using Python. A total of 2,000 reviews were gathered. After removing duplicate and irrelevant reviews through data cleaning, the final dataset consisted of 1,921 reviews. This study will classify the data into positive and negative sentiments using machine learning.

In the initial observation, the reviews showed a mix of feedback. Many users found the app beginner-friendly, while others raised concerns about its stability and compatibility.

This dataset can be useful for researchers conducting sentiment analysis in the cryptocurrency investment industry in Indonesia, helping them understand user experiences and identify areas for improvement. Additionally, it can serve as training data for machine learning models in sentiment classification. By analyzing user feedback, this dataset can serve as a foundation for investment apps to enhance application performance and preserve essential features while refining areas that need improvement in the development of cryptocurrency investment applications in Indonesia.

本数据集收录了针对投资应用Nanovest的用户评论，该应用面向AS股票、黄金及加密货币（cryptocurrency）提供投资服务。数据采集自Google Play商店中Nanovest应用的用户评论，所有评论均以印尼语撰写，完整反映了用户对该应用功能、安全性及运行性能的使用体验与评价。本研究通过分析该评论数据，旨在明确正负向评论的占比，并挖掘用户高频提及的核心评价维度，其研究结果可为印尼境内加密货币投资平台的服务质量与应用性能优化提供切实参考。

本数据集通过Python编写的网络爬虫（web scraping）采集，初始共获取2000条用户评论。经数据清洗步骤剔除重复与无关评论后，最终有效数据集包含1921条评论。本研究将采用机器学习（machine learning）方法对该数据集开展正负向情感分类（sentiment classification）任务。

初步观测表明，评论反馈呈现两极分化特征：不少用户认为该应用对新手友好，但也有部分用户对其运行稳定性与兼容性提出了担忧。

该数据集可用于印尼加密货币投资领域的情感分析（sentiment analysis）研究，助力研究者洞悉用户体验并定位待优化方向；此外，其亦可作为情感分类任务中机器学习模型的训练数据集。通过剖析用户反馈，本数据集可为印尼加密货币投资应用的研发提供核心参考依据：帮助应用在保留核心功能的前提下，优化待改进环节，进而全面提升应用整体性能。