← heapsort-ai

sentiment analysis

16 items

RESEARCHarXiv CS.CL·4/20/2026

Consistency Analysis of Sentiment Predictions using Syntactic & Semantic Context Assessment Summarization (SSAS)

This paper introduces the Syntactic & Semantic Context Assessment Summarization (SSAS) framework to address the inconsistency of LLM-based sentiment predictions, a challenge for reliable enterprise analytics. SSAS functions as a sophisticated data pre-processing tool, employing hierarchical classification and iterative summarization to establish high-signal, sentiment-dense context for more stable and reliable business decisions.

33
RESEARCHarXiv CS.CL·4/16/2026

A Multi-Model Approach to English-Bangla Sentiment Classification of Government Mobile Banking App Reviews

This study classifies sentiment in English and Bangla reviews of Bangladeshi government mobile banking apps, using a hybrid labeling approach for 5,652 reviews. It found that traditional machine learning models like Random Forest and Linear SVM significantly outperformed fine-tuned XLM-RoBERTa for this specific task.

31
ARTICLEDEV.to AI·7d ago

Mining for Gold: Using AI Sentiment Triage to Identify Your DTC Super-Fans

This article explores how AI sentiment triage can help DTC brands identify and prioritize their most valuable customers, the "super-fans". By classifying support interactions based on advocacy potential, companies can optimize engagement and leverage the high lifetime value (LTV) and referrals of these customers. Tools like OpenAI's API (GPT-4) can automate this analysis, turning high-volume tickets into a growth-driven strategy.

28
RESEARCHarXiv CS.CL·21d ago

Beyond Sentiment Classification: A Generative Framework for Emotion Intensity Evaluation in Text

This work introduces a novel approach to emotion modeling, shifting from discrete classification to continuous emotion intensity evaluation in text. The authors constructed a dataset of emotional intensity scores and fine-tuned generative language models to output continuous values from 0-100, outperforming classification baselines and demonstrating generalization capabilities.

28
RESEARCHarXiv CS.CL·19d ago

Leveraging Large Language Models for Sentiment Analysis: Multi-Modal Analysis of Decentraland's MANA Token

This study explores the integration of Decentraland's Discord community sentiment analysis, using a BERT-based large language model, with multi-modal financial data to predict the MANA token price. Results indicate that a multi-modal model, incorporating sentiment, trading volume, and market capitalization, significantly outperforms a price-only prediction baseline.

27