Recent Research on the Development of Emotion Analysis 
Technology in Text Mining

Xinzhu Li

doi:10.70711/frim.v3i6.6663

Recent Research on the Development of Emotion Analysis Technology in Text Mining

Xinzhu Li

Abstract

This paper aims to study the latest progress of emotion analysis technology in text mining, and discuss the application and performance of different models (LSTM, SVM, BERT) in emotion classification tasks. Three models were analyzed by IMDB, including precision,
recall, F1 value, and AUC value. The experimental results showed that the BERT model was better than LSTM and SVM in all indexes, with
88.2% accuracy and an AUC value of 0.93, indicating a significant advantage in the emotion analysis task. The study in this paper provides a
reference for future optimization of emotion analysis methods, especially the potential of BERT model in emotion analysis.

Keywords

Text mining; Sentiment analysis; BERT; LSTM

Full Text:

PDF

Included Database

References

[1] Lyu W, Chen J, Qin X, et al. Word Vector Representation of Latin Cuengh Based on Root Feature Enhancement[J]. Applied Sciences,

2024, 15(1):211-211.

[2] Jin X, Wei F, Kandala S S, et al. Time series forecasting of Valley fever infection in Maricopa County, AZ using LSTM.[J]. Lancet regional health. Americas, 2025, 15(11):21-23.

[3] Crossley A S. Developing Linguistic Constructs of Text Readability Using Natural Language Processing[J]. Scientific Studies of Reading, 2025, 29(2):138-160.

[4] Wan H, Fang Y, Hu M, et al. Interpretable Machine-Learning and Big Data Mining to Predict the CO2 Separation in Polymer-MOF

Mixed Matrix Membranes.[J]. Advanced science (Weinheim, Baden-Wurttemberg, Germany), 2025, 20(15):15-17.

[5] Monaco S, Cortellino A, Massari G M, et al. Using Emotional Text Mining to Explore the Cultural Representation of Organ Donation in

Spanish and Italian Culture[J]. Culture & Psychology, 2025, 31(1):259-276.

DOI: http://dx.doi.org/10.70711/frim.v3i6.6663

Refbacks

There are currently no refbacks.