<![CDATA[In the rapidly advancing era of digital transformation, where textual data abounds from various online sources such as social media, forums, and product reviews, sentiment analysis has become a critical component in understanding public opinions and consumer behavior. Sentiment analysis employs machine learning, natural language processing, and computational linguistics to comprehend the feelings and opinions of others. The machine learning algorithms investigated in this paper include K-Nearest Neighbor (K-NN), Support Vector Machine (SVM), Naive Bayes, ID3, and C4.5. The sentiment analysis process requires significant computational resources to handle the complexity and scale of data. This research aims to examine the differences in resource usage among these algorithms and determine which algorithm is best suited for sentiment analysis in this context. The research methodology employed is quantitative, focusing on the collection and numerical analysis of datasets. Testing is conducted using the Anaconda Library in the Python programming language to measure the usage of Resident Set Size (RSS), Virtual Memory Size (VMS), execution time, and the accuracy of each algorithm. The test results indicate that the Support Vector Machine (SVM) algorithm with an accuracy rate of 96% and the Naive Bayes algorithm with an accuracy rate of 97% are the best choices for use in the context of sentiment analysis. When considering the context of Resident Set Size (RSS) and Virtual Memory Size (VMS) usage in a single execution, ID3 is the algorithm with the smallest resource usage, with an accuracy rate of 92%. The average resources used by ID3 are 8.318.566,4 bytes for Resident Set Size (RSS) and 7.965.900,8 bytes for Virtual Memory Size (VMS) with an execution time of 2,619 seconds.]]>
Copyrights © 2024
