# Leveraging Twitter Data for Sentiment Analysis of Transit User Feedback: An NLP Framework

arXiv: https://arxiv.org/abs/2310.07086 · HTML: https://mlsanity.com/paper/2310.07086/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Builds on 17 of its 83 references that are in the map.
- First posted on arXiv 11 Oct 2023.

## Builds on (most cited first, 17 of 17)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [XGBoost: A Scalable Tree Boosting System](https://mlsanity.com/paper/1603.02754/) (2016)
- [Deep contextualized word representations](https://mlsanity.com/paper/1802.05365/) (2018)
- [Convolutional Neural Networks for Sentence Classification](https://mlsanity.com/paper/1408.5882/) (2014)
- [Pre-trained Models for Natural Language Processing: A Survey](https://mlsanity.com/paper/2003.08271/) (2020)
- [Few-shot Learning with Multilingual Language Models](https://mlsanity.com/paper/2112.10668/) (2021)
- [Supervised Contrastive Learning for Pre-trained Language Model Fine-tuning](https://mlsanity.com/paper/2011.01403/) (2020)
- [An Embarrassingly Simple Approach for Transfer Learning from Pretrained Language Models](https://mlsanity.com/paper/1902.10547/) (2019)
- [Better Document-level Sentiment Analysis from RST Discourse Parsing](https://mlsanity.com/paper/1509.01599/) (2015)
- [Just Another Day on Twitter: A Complete 24 Hours of Twitter Data](https://mlsanity.com/paper/2301.11429/) (2023)
- [Detecting Topic and Sentiment Dynamics Due to COVID-19 Pandemic Using Social Media](https://mlsanity.com/paper/2007.02304/) (2020)
- [A Generative Language Model for Few-shot Aspect-Based Sentiment Analysis](https://mlsanity.com/paper/2204.05356/) (2022)
- [Hostility Detection in Hindi leveraging Pre-Trained Language Models](https://mlsanity.com/paper/2101.05494/) (2021)
- [Mining Social Media for Open Innovation in Transportation Systems](https://mlsanity.com/paper/1610.09894/) (2016)
- [Fuzzy Ontology-Based Sentiment Analysis of Transportation and City Feature Reviews for Safe Traveling](https://mlsanity.com/paper/1701.05334/) (2017)

## Built on by (most cited first, 1 of 1)

- [Generative AI in Transportation Planning: A Survey](https://mlsanity.com/paper/2503.07158/) (2025)

## Abstract (arXiv)

Traditional methods of collecting user feedback through transit surveys are often time-consuming, resource intensive, and costly. In this paper, we propose a novel NLP-based framework that harnesses the vast, abundant, and inexpensive data available on social media platforms like Twitter to understand users' perceptions of various service issues. Twitter, being a microblogging platform, hosts a wealth of real-time user-generated content that often includes valuable feedback and opinions on various products, services, and experiences. The proposed framework streamlines the process of gathering and analyzing user feedback without the need for costly and time-consuming user feedback surveys using two techniques. First, it utilizes few-shot learning for tweet classification within predefined categories, allowing effective identification of the issues described in tweets. It then employs a lexicon-based sentiment analysis model to assess the intensity and polarity of the tweet sentiments, distinguishing between positive, negative, and neutral tweets. The effectiveness of the framework was validated on a subset of manually labeled Twitter data and was applied to the NYC subway system as a case study. The framework accurately classifies tweets into predefined categories related to safety, reliability, and maintenance of the subway system and effectively measured sentiment intensities within each category. The general findings were corroborated through a comparison with an agency-run customer survey conducted in the same year. The findings highlight the effectiveness of the proposed framework in gauging user feedback through inexpensive social media data to understand the pain points of the transit system and plan for targeted improvements.