In today’s rapidly expanding e-commerce environment, the massive volume of product reviews makes it crucial to summarize user opinions in a way that is both comprehensible and practically applicable. This research presents a system for analyzing product reviews using Aspect-Based Sentiment Analysis (ABSA), a Natural Language Processing (NLP) technique that identifies key aspects of a review (such as shipping, product quality, and packaging) and evaluates the sentiment (positive, negative, or neutral) associated with each aspect, allowing both consumers and merchants to gain more efficient access to in-depth insights. This project focuses on developing AI for Thai-language ABSA by utilizing WangchanBERTa, a model trained on Thai data, and comparing it with various standard approaches such as TF-IDF + Logistic Regression, Word2Vec + BiLSTM, and Multilingual BERT (mBERT/XLM-R) to assess their performance in terms of accuracy, speed, and resource usage. Additionally, a dashboard visualization is provided to help users quickly grasp review trends. The expected outcome is to create an AI tool that can be practically employed in the e-commerce industry, enabling consumers to make easier purchasing decisions and assisting merchants in effectively improving their products and services.
ปัจจุบันการซื้อขายสินค้าออนไลน์เติบโตขึ้นอย่างรวดเร็ว ทำให้ข้อมูลรีวิวสินค้าจากผู้บริโภคมีปริมาณเพิ่มขึ้นเป็นจำนวนมาก รีวิวเหล่านี้มีบทบาทสำคัญในการตัดสินใจซื้อของลูกค้าและการปรับปรุงคุณภาพสินค้าของร้านค้า อย่างไรก็ตาม ปริมาณข้อมูลที่มากเกินไปและความหลากหลายของรูปแบบการแสดงความคิดเห็นทำให้การสรุปและวิเคราะห์ข้อมูลเหล่านี้เป็นไปได้ยาก ผู้บริโภคต้องใช้เวลามากในการอ่านรีวิวจำนวนมากเพื่อสรุปแนวโน้มความคิดเห็น ในขณะที่ร้านค้าประสบปัญหาในการวิเคราะห์ข้อมูลรีวิวเพื่อปรับปรุงผลิตภัณฑ์และบริการ เพื่อแก้ไขปัญหานี้ งานวิจัยนี้นำเสนอการประยุกต์ใช้ Aspect-Based Sentiment Analysis (ABSA) ซึ่งเป็นเทคนิคใน Natural Language Processing (NLP) ที่สามารถแยกแยะ แง่มุมสำคัญของรีวิวสินค้า (Aspects) และวิเคราะห์ อารมณ์ของแต่ละแง่มุม (Sentiments) โดยอัตโนมัติ การนำเทคนิคนี้มาใช้จะช่วยให้ผู้บริโภคสามารถรับข้อมูลเชิงลึกจากรีวิวได้ง่ายขึ้น และช่วยให้ร้านค้าสามารถใช้ข้อมูลรีวิวเพื่อปรับปรุงสินค้าและบริการอย่างมีประสิทธิภาพ โครงงานนี้ยังมุ่งเน้นการศึกษาว่า แนวทาง AI แบบใดมีประสิทธิภาพสูงสุดในการทำ ABSA สำหรับภาษาไทย โดยเปรียบเทียบ วิธีการประมวลผลภาษาธรรมชาติแบบดั้งเดิม กับเทคนิคการเรียนรู้เชิงลึกที่ทันสมัย เพื่อให้ได้แนวทางที่เหมาะสมที่สุด พร้อมทั้งพัฒนา Dashboard Visualization ที่ช่วยให้ข้อมูลรีวิวถูกนำเสนอในรูปแบบที่เข้าใจง่ายและสามารถนำไปใช้งานจริงในอุตสาหกรรมอีคอมเมิร์ซ
คณะวิศวกรรมศาสตร์
In Thailand, the quantity of old tires has been increasing annually, posing a significant environmental challenge due to their non-biodegradable material. However, old tires contain an internal porous structure, which suggests their potential application as sound-absorbing materials. Porosity is a key characteristic that enables materials to trap sound waves, making them effective for noise reduction. Therefore, this study aims to investigate and develop sound-absorbing materials from old tire rubber powder. The methodology involved mixing old tire powder with fresh latex at a ratio of 1:2, followed by drying at a temperature of 120°C for four hours. Subsequently, the physical properties influencing sound absorption, including density, porosity, and water absorption, were analyzed. The results indicated that the sound-absorbing material produced from old tire rubber powder showed a density of 0.96 g/cm³, a porosity value of 0.45, and a water absorption of 11.03%. Therefore, the findings suggest that old tire rubber powder has the potential to be effectively utilized as a sound-absorbing material.
คณะครุศาสตร์อุตสาหกรรมและเทคโนโลยี
This aimed to 1) develop an effective augmented reality (AR) media integrated with the metaverse to enhance English phonics and communication skills. 2) To evaluate English pronunciation skills using augmented reality media integrated with the metaverse, and 3) To assess English communication skills through interactions within the metaverse. The sample group comprised 120 Grade 4 students from two classrooms in the first semester of the 2024 academic year, selected through cluster random sampling and divided into experimental and control groups. The research instruments included AR media sets, media quality assessment forms, phonics tests, and English communication skills assessment forms, administered before and after the learning intervention. Data analysis employed mean (x ̅), standard deviation (S), t-tests for independent samples, and one-way analysis of variance (Multivariate Analysis of Variance: One-Way MANOVA) to compare mean score differences between the experimental and control groups. Results indicated that the overall quality of the AR media kit with the metaverse was rated at a very high level (x ̅= 4.80, S.D. = 0.12). Evaluating specific aspects showed that the content quality was at the highest level (x ̅= 4.92, S.D. = 0.07), while the media production technique also rated highly (x ̅ = 4.70, S.D. = 0.17). Furthermore, the English pronunciation and communication skills of the Grade 4 students using the AR media with the metaverse were significantly higher after the intervention compared to before, the overall quality of the AR media integrated with the metaverse was rated at the highest level (x ̅= 4.80, S.D. = 0.12). For individual aspects, content quality was rated at the highest level (x ̅= 4.92, S.D. = 0.07), and media production techniques were also rated at the highest level (x ̅ = 4.70, S.D. = 0.17). Comparing the mean scores of English pronunciation and communication skills between the two groups, it was found that the experimental group using AR media integrated with the metaverse demonstrated significantly higher English pronunciation skills than the control group (F(1, 89) = 3261.422, p = 0.001, Partial η² = 0.98). Additionally, the experimental group exhibited significantly higher English communication skills than the control group (F(1, 89) = 4239.365, p = 0.001, Partial η² = 0.98). These results aligned with the research hypothesis that "Grade 4 students’ English pronunciation and communication skills post-learning with AR media integrated with the metaverse would significantly improve compared to their pre-learning levels at the .05 level of significance.
คณะสถาปัตยกรรม ศิลปะและการออกแบบ
The thesis artwork titled “The Red Mist” presents a narrative adapted from a short story of the same name by Assistant Professor Chatnarong Wisutkul in 2003. The story is set in a future world where people's greed and selfishness have led to a war, forcing them to rely on "breathing machines" to survive in the "red toxic mist." Phakin, a 15-year-old boy, embarks on a journey with a group of refugees. As they pass through abandoned cities, they encounter a boy without a breathing machine who has recently lost his father. Phakin decides to help him, despite objections from others. The boy tries to end his life by shutting off his breathing machine, and when Phakin intervenes to save him, he collapses from inhaling the toxic air. Witnessing Phakin's selfless act, the others are moved and join forces to save both of them. Phakin demonstrates that in difficult times, humans must cooperate and help each other rather than being divided and selfish.