Sentiment Analysis on Machine LearningBased on HP Taccola Printer Customer Sentiment Data10/29/2021 Author : Spencer Chang Mountain View High School, Vancouver, Washington, USA Abstract The Trillium printer coming out for the next generation of HP printers comes from the design of the Taccola printer, which had notably bad reviews. In order to make a better printer, HP wanted to capitalize on the weak points of the Taccola printer, keep the strong points, and design Trillium with the best of both worlds. 17000 customer comments from the last 3 months were analyzed to find general sentiment on the product. Using Python machine learning, the comments were run though sentiment models in order to deliver a minimum of 80% of sentiments correct. Using Sentence-Bert, SentenceTransformer, and spatial distance, a 90% accuracy rate was achieved. In conclusion, the main weakness of the Taccola needs to change its app in order to function smoother. ![]()
0 Comments
Leave a Reply. |
October 2021 IssueCategoriesArchives |