# SentimentArcs: A Novel Method for Self-Supervised Sentiment Analysis of Time Series Shows SOTA Transformers Can Struggle Finding Narrative Arcs

arXiv: https://arxiv.org/abs/2110.09454 · HTML: https://mlsanity.com/paper/2110.09454/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Builds on 21 of its 104 references that are in the map.
- First posted on arXiv 18 Oct 2021.

## Builds on (most cited first, 21 of 21)

- [BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding](https://mlsanity.com/paper/1810.04805/) (2018)
- [XGBoost: A Scalable Tree Boosting System](https://mlsanity.com/paper/1603.02754/) (2016)
- [Efficient Estimation of Word Representations in Vector Space](https://mlsanity.com/paper/1301.3781/) (2013)
- [Stanza: A Python Natural Language Processing Toolkit for Many Human Languages](https://mlsanity.com/paper/2003.07082/) (2020)
- [Ensemble deep learning: A review](https://mlsanity.com/paper/2104.02395/) (2021)
- [Crowdsourcing a Word-Emotion Association Lexicon](https://mlsanity.com/paper/1308.6297/) (2013)
- [Challenges in Deploying Machine Learning: a Survey of Case Studies](https://mlsanity.com/paper/2011.09926/) (2020)
- [sktime: A Unified Interface for Machine Learning with Time Series](https://mlsanity.com/paper/1909.07872/) (2019)
- [The emotional arcs of stories are dominated by six basic shapes](https://mlsanity.com/paper/1606.07772/) (2016)
- [UDALM: Unsupervised Domain Adaptation through Language Modeling](https://mlsanity.com/paper/2104.07078/) (2021)
- [Revisiting LSTM Networks for Semi-Supervised Text Classification via Mixed Objective Function](https://mlsanity.com/paper/2009.04007/) (2020)
- [BB_twtr at SemEval-2017 Task 4: Twitter Sentiment Analysis with CNNs and LSTMs](https://mlsanity.com/paper/1704.06125/) (2017)
- [SemEval-2016 Task 4: Sentiment Analysis in Twitter](https://mlsanity.com/paper/1912.01973/) (2019)
- [Comparing and Combining Sentiment Analysis Methods](https://mlsanity.com/paper/1406.0032/) (2014)
- [Long Short-Term Memory with Dynamic Skip Connections](https://mlsanity.com/paper/1811.03873/) (2018)
- [Leveraging Automated Machine Learning for Text Classification: Evaluation of AutoML Tools and Comparison with Human Performance](https://mlsanity.com/paper/2012.03575/) (2020)
- [Improved Sentence Modeling using Suffix Bidirectional LSTM](https://mlsanity.com/paper/1805.07340/) (2018)
- [Anger is More Influential Than Joy: Sentiment Correlation in Weibo](https://mlsanity.com/paper/1309.2402/) (2013)
- [Amobee at SemEval-2017 Task 4: Deep Learning System for Sentiment Detection on Twitter](https://mlsanity.com/paper/1705.01306/) (2017)
- [Can Sentiment Analysis Reveal Structure in a Plotless Novel?](https://mlsanity.com/paper/1910.01441/) (2019)
- [Predicting market inflation expectations with news topics and sentiment](https://mlsanity.com/paper/2107.07155/) (2021)

## Built on by (most cited first, 4 of 4)

- [NarraBench: A Comprehensive Framework for Narrative Benchmarking](https://mlsanity.com/paper/2510.09869/) (2025)
- [Experimenting with UD Adaptation of an Unsupervised Rule-based Approach for Sentiment Analysis of Mexican Tourist Texts](https://mlsanity.com/paper/2309.05312/) (2023)
- [The AI Fiction Paradox](https://mlsanity.com/paper/2603.13545/) (2026)
- [LLMs vs. Traditional Sentiment Tools in Psychology: An Evaluation on Belgian-Dutch Narratives](https://mlsanity.com/paper/2511.07641/) (2025)

## Abstract (arXiv)

SOTA Transformer and DNN short text sentiment classifiers report over 97% accuracy on narrow domains like IMDB movie reviews. Real-world performance is significantly lower because traditional models overfit benchmarks and generalize poorly to different or more open domain texts. This paper introduces SentimentArcs, a new self-supervised time series sentiment analysis methodology that addresses the two main limitations of traditional supervised sentiment analysis: limited labeled training datasets and poor generalization. A large ensemble of diverse models provides a synthetic ground truth for self-supervised learning. Novel metrics jointly optimize an exhaustive search across every possible corpus:model combination. The joint optimization over both the corpus and model solves the generalization problem. Simple visualizations exploit the temporal structure in narratives so domain experts can quickly spot trends, identify key features, and note anomalies over hundreds of arcs and millions of data points. To our knowledge, this is the first self-supervised method for time series sentiment analysis and the largest survey directly comparing real-world model performance on long-form narratives.