---
title: "Citius — paper by Gamallo and Garcia (2014)"
description: "The authors build a family of Naive Bayes classifiers in Perl. 12 references, cited by 3 papers here."
authors: ["Pablo Gamallo", "Marcos Garcia"]
year: 2014
venue: "Proceedings of the 8th International Workshop on Semantic Evaluation (SemEval 2014)"
source: "https://aclanthology.org/S14-2026/"
doi: "10.3115/v1/s14-2026"
licence: "CC BY-NC-SA 3.0 (ACL Anthology, before 2016)"
licence_url: "https://creativecommons.org/licenses/by-nc-sa/3.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/gamallo-2014-citius-naive-bayes-strategy-sentiment/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/gamallo-2014-citius-naive-bayes-strategy-sentiment/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2014](https://md.research.elamur.ai/free/en/papers/2014/page/2/) / Citius: A Naive-Bayes Strategy for Sentiment Analysis on English Tweets

# Citius: A Naive-Bayes Strategy for Sentiment Analysis on English Tweets

**Authors:** [Pablo Gamallo](https://md.research.elamur.ai/free/en/authors/pablo-gamallo/) and [Marcos Garcia](https://md.research.elamur.ai/free/en/authors/marcos-garcia/) · **Year:** 2014 · **Venue:** [Proceedings of the 8th International Workshop on Semantic Evaluation (SemEval 2014)](https://md.research.elamur.ai/free/en/venues/semeval/) ([SemEval 2014](https://md.research.elamur.ai/free/en/venues/semeval/2014/)) · **Publisher:** [Association for Computational Linguistics](https://md.research.elamur.ai/free/en/publishers/association-for-computational-linguistics/) · **Licence:** [CC BY-NC-SA 3.0 (ACL Anthology, before 2016)](https://creativecommons.org/licenses/by-nc-sa/3.0/)

**Links:** [Publisher's page](https://aclanthology.org/S14-2026/) · [PDF](https://aclanthology.org/S14-2026.pdf) · [DOI 10.3115/v1/s14-2026](https://doi.org/10.3115/v1/s14-2026) · [OpenAlex](https://openalex.org/W2169816422)

**Citations:** 132 citations (OpenAlex, 2026-10-14) · 3 in our library, of the 149,439 papers whose bibliographies we read

## Summary
The paper addresses SemEval-2014 Task 9, subtask B: deciding whether an English tweet is positive, negative or neutral. The authors build a family of Naive Bayes classifiers in Perl. Before lemmatization, tweets are normalized: URLs, usernames and hashtags are removed, stretched letters are collapsed, and emoticons and interjections become sentiment words. Features are content-word lemma unigrams, POS-pattern multiwords, a merged polarity lexicon and negation handling that depends on part of speech. The central design choice splits the problem in two. A binary classifier is trained only on positive and negative tweets, and a simple lexicon test labels a tweet neutral when it contains no polarity lemma. This split beat a direct three-class model. Adding the lexicon and multiwords pushed F-score from 0.49 to 0.63. The best submission ranked 26th of 53 systems, which is mid-table. The work shows that a cheap, transparent pipeline built from linguistically motivated features can reach average competitive performance.

## Key points
- Data: 6,408 SemEval training tweets (tweeti-b.dist.tsv), with Neutral, Objective and Neutral-or-Objective merged into one Neutral label. "Unconstrained" runs also used 5,050 external positive and negative tweets (Narr et al., 2012). Development testing used tweeti-b.devel.tsv.
- The polarity lexicon has 10,850 entries merged from AFINN-111 (1,520 lemmas), Hedonometer (top 300 positive), the Hu&Liu list (5,695) and SentiWordNet-3.0 (6,600 strongest lemmas).
- Multiwords came from POS patterns such as NOUN-ADJ, ADJ-NOUN, NOUN-NOUN, VERB-NOUN, NOUN-PRP-NOUN and VERB-PRP-NOUN, extracted with the authors' GPL tool GaleXtra.
- Development F-scores were:

  | System | F-score |
  |---|---|
  | CONSTR-BASE | .49 |
  | CONSTR-BASE-LEX | .56 |
  | CONSTR-BIN-LEX | .57 |
  | CONSTR-BIN-LEX-MW | .61 |
  | UNCONSTR-BIN-LEX | .58 |
  | UNCONSTR-BIN-LEX-MW | .63 |

- On the official Tweets2014 test set, the constrained system scored 0.62 and the unconstrained one 0.63. A Spanish variant placed 3rd of 13 at TASS-2013.

## Interesting ideas
- Each lexicon entry is turned into a synthetic one-lemma training example that carries the entry's polarity, which lets a statistical model absorb a hand-built lexicon without any architecture change.
- Instead of one three-way decision, a lexicon gate first decides whether there is any polarity, and a two-class model then picks the direction; the clean split outperformed the joint classifier.
- Negation scope is chosen by the negator's category: "not" flips only nearby verbs and adjectives, while "no" and "none" flip only nouns, all within a short window.
- POS-tag patterns pick out linguistically motivated multiword units, mostly syntactically linked pairs, and adding them gave the single biggest gain in the ablation.
- Evaluation is run as a ladder of systems that each add one component, plus a split between constrained and unconstrained training data, so each component's contribution is isolated.

## For Mova

How these ideas could serve Mova, our small open language model — its files, stages and tests:

- **I1** (learn): Dictionary knowledge such as VerbNet classes or WordNet senses could be injected as minimal synthetic training items, giving rare words coverage the treebank or silver data lacks.
- **I2** (frames): Speech-act or irony detection could first ask a cheap, explainable "is anything evaluative here?" question before running a finer classifier, which keeps every decision traceable.
- **I3** (frames): The same idea maps naturally onto UD arcs (advmod vs det negation attached to a head), so polarity reversal in frames can follow the tree rather than word distance.
- **I4** (tokens): Pattern-based multiword candidates could feed the lemmatizer/tagger as compound units, so that later stages treat them as one concept.
- **I5** (learn): Evaluation reports could adopt the same incremental-ablation layout and keep runs on in-house data separate from runs with external data, so data effects are not confused with method effects.

*Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.*

## Institutions

- [Universidade de Santiago de Compostela](https://md.research.elamur.ai/free/en/universities/universidade-de-santiago-de-compostela/) — University, Santiago de Compostela, Spain · 13 works · 364 citations

📚 Shelves (2): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 18,614 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 17,965 works

🧭 Topics (2): [English](https://md.research.elamur.ai/free/en/topics/english/) · 2,379 works · [😊 Sentiment & emotion](https://md.research.elamur.ai/free/en/topics/sentiment-emotion/) · 1,035 works

## Cite

Cite as: Pablo Gamallo and Marcos Garcia. 2014. Citius: A Naive-Bayes Strategy for Sentiment Analysis on English Tweets. Proceedings of the 8th International Workshop on Semantic Evaluation (SemEval 2014).

The paper itself, not this page. BibTeX from the library's record of the work.

```bibtex
@inproceedings{gamallo2014citius,
  title = {{Citius: A Naive-Bayes Strategy for Sentiment Analysis on English Tweets}},
  author = {Pablo Gamallo and Marcos Garcia},
  booktitle = {Proceedings of the 8th International Workshop on Semantic Evaluation (SemEval 2014)},
  year = {2014},
  doi = {10.3115/v1/s14-2026},
  url = {https://doi.org/10.3115/v1/s14-2026}
}
```

- [← previous A Comparison of the Events and Relations Across ACE, ERE, TAC-KBP, and FrameNet Annotation Standards](https://md.research.elamur.ai/free/en/papers/aguilar-2014-comparison-events-relations-across-ace/)
- [next → Major Life Event Extraction from Twitter based on Congratulations/Condolences Speech Acts](https://md.research.elamur.ai/free/en/papers/li-2014-major-life-event-extraction-twitter/)
- [↑ back to papers of 2014](https://md.research.elamur.ai/free/en/papers/2014/page/2/)

## References

12 references, 2 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Antonio Fernández Anta, Luis Núñez Chiroque, Philippe Morere, and Agustı́n Santos. 2013. Sentiment Analysis and Topic Detection of Spanish Tweets: A Comparative Study of NLP Techniques. Procesamiento del Lenguaje Natural, 50:45–52. [find](https://scholar.google.com/scholar?q=Antonio+Ferna%CC%81ndez+Anta%2C+Luis+Nu%CC%81n%CC%83ez+Chiroque%2C+Philippe+Morere%2C+and+Agust%C4%B1%CC%81n+Santos%2E+2013%2E+Sentiment+Analysis+and+Topic+Detection+of+Spanish+Tweets%3A+A+Comparative+Study+of+NLP+Techniques%2E+Procesamien)
2. Stefano Baccianella, Andrea Esuli, and Fabrizio Sebastiani. 2010. [SentiWordNet 3.0: An Enhanced Lexical Resource for Sentiment Analysis and Opinion Mining](https://md.research.elamur.ai/free/en/papers/baccianella-2010-sentiwordnet-3-0-enhanced-lexical/). In Human Language Technology Conference - North American chapter of the Association for Computational Linguistics, pages 2200–2204.
3. Peter Sheridan Dodds, Kameron Decker Harris, Isabel M. Kloumann, Catherine A. Bliss, and Christopher M. Danforth. 2011. Temporal patterns of happiness and information in a global social network: Hedonometrics and Twitter. PLoS ONE, 6(12):e26752. [find](https://scholar.google.com/scholar?q=Peter+Sheridan+Dodds%2C+Kameron+Decker+Harris%2C+Isabel+M%2E+Kloumann%2C+Catherine+A%2E+Bliss%2C+and+Christopher+M%2E+Danforth%2E+2011%2E+Temporal+patterns+of+happiness+and+information+in+a+global+social+network%3A+Hedon)
4. Pablo Gamallo, Marcos Garcia, and Santiago Fernández-Lanza. 2013. TASS: A Naive-Bayes strategy for sentiment analysis on Spanish tweets. In Workshop on Sentiment Analysis at SEPLN (TASS2013), pages 126–132, Madrid, Spain. [find](https://scholar.google.com/scholar?q=Pablo+Gamallo%2C+Marcos+Garcia%2C+and+Santiago+Ferna%CC%81ndez%2DLanza%2E+2013%2E+TASS%3A+A+Naive%2DBayes+strategy+for+sentiment+analysis+on+Spanish+tweets%2E+In+Workshop+on+Sentiment+Analysis+at+SEPLN+%28TASS2013%29%2C+pages+1)
5. Alec Go, Richa Bhayani, and Lei Huang. 2009. Twitter sentiment classification using distant supervision. In CS224N Technical report. Standford. [find](https://scholar.google.com/scholar?q=Alec+Go%2C+Richa+Bhayani%2C+and+Lei+Huang%2E+2009%2E+Twitter+sentiment+classification+using+distant+supervision%2E+In+CS224N+Technical+report%2E+Standford%2E)
6. Bing Liu, Minqing Hu, and Junsheng Cheng. 2005. Opinion observer: Analyzing and comparing opinions on the web. In 14th International World Wide Web conference (WWW-2005), pages 342–351, New York, NY, USA. [find](https://scholar.google.com/scholar?q=Bing+Liu%2C+Minqing+Hu%2C+and+Junsheng+Cheng%2E+2005%2E+Opinion+observer%3A+Analyzing+and+comparing+opinions+on+the+web%2E+In+14th+International+World+Wide+Web+conference+%28WWW%2D2005%29%2C+pages+342%E2%80%93351%2C+New+York%2C+NY%2C+)
7. Chris Manning, Prabhakar Raghadvan, and Hinrich Schütze. 2008. Introduction to Information Retrieval. Cambridge University Press, Cambridge, MA, USA. [find](https://scholar.google.com/scholar?q=Chris+Manning%2C+Prabhakar+Raghadvan%2C+and+Hinrich+Schu%CC%88tze%2E+2008%2E+Introduction+to+Information+Retrieval%2E+Cambridge+University+Press%2C+Cambridge%2C+MA%2C+USA%2E)
8. Mario Barcala and Eva Domı́nguez and Pablo Gamallo and Marisol López and Eduardo Moscoso and Guillermo Rojo and Paula Santalla and Susana Sotelo. 2007. A Corpus and Lexical Resources for Multi-word Terminology Extraction in the Field of Economy. In 3rd Language & Technology Conference (LeTC’2007), pages 355–359, Poznan, Poland. [find](https://scholar.google.com/scholar?q=Mario+Barcala+and+Eva+Dom%C4%B1%CC%81nguez+and+Pablo+Gamallo+and+Marisol+Lo%CC%81pez+and+Eduardo+Moscoso+and+Guillermo+Rojo+and+Paula+Santalla+and+Susana+Sotelo%2E+2007%2E+A+Corpus+and+Lexical+Resources+for+Multi%2Dword+T)
9. Sascha Narr, Michael Hulfenhaus, and Sahin Albayrak. 2012. Language-Independent Twitter Sentiment Analysis. In Knowledge Discovery and Machine Learning (KDML), LWA, pages 12–14, Dortmund, Germany. [find](https://scholar.google.com/scholar?q=Sascha+Narr%2C+Michael+Hulfenhaus%2C+and+Sahin+Albayrak%2E+2012%2E+Language%2DIndependent+Twitter+Sentiment+Analysis%2E+In+Knowledge+Discovery+and+Machine+Learning+%28KDML%29%2C+LWA%2C+pages+12%E2%80%9314%2C+Dortmund%2C+Germany%2E)
10. Alexander Pak and Patrick Paroubek. 2010. [Twitter as a Corpus for Sentiment Analysis and Opinion Mining](https://md.research.elamur.ai/free/en/papers/pak-2010-twitter-as-corpus-sentiment-analysis/). In LREC-2010, Valletta, Malta.
11. Julio Villena-Román, Sara Lana, Eugeinio Martı́nez-Cámara, and Juan Carlos González-Cristóbal. 2013. TASS - Workshop on Sentiment Analysis at SEPLN. Procesamiento del Lenguaje Natural, 50:37–44. [find](https://scholar.google.com/scholar?q=Julio+Villena%2DRoma%CC%81n%2C+Sara+Lana%2C+Eugeinio+Mart%C4%B1%CC%81nez%2DCa%CC%81mara%2C+and+Juan+Carlos+Gonza%CC%81lez%2DCristo%CC%81bal%2E+2013%2E+TASS+%2D+Workshop+on+Sentiment+Analysis+at+SEPLN%2E+Procesamiento+del+Lenguaje+Natural%2C+50%3A37%E2%80%9344%2E)
12. Kiduk Yang. 2008. WIDIT in TREC 2008 blog track: Leveraging Multiple Sources of Opinion Evidence. In The Seventeenth Text Retrieval Conference (TREC-2008), Gaithersburg, Maryland, USA. [find](https://scholar.google.com/scholar?q=Kiduk+Yang%2E+2008%2E+WIDIT+in+TREC+2008+blog+track%3A+Leveraging+Multiple+Sources+of+Opinion+Evidence%2E+In+The+Seventeenth+Text+Retrieval+Conference+%28TREC%2D2008%29%2C+Gaithersburg%2C+Maryland%2C+USA%2E)

*Updated 2 November 2026*