Lexicon

CRiSOL

Resource type:

Lexicon

Description:

CRiSOL is the result of the combination of two linguistic resources for Sentiment Analysis. One of those resources is iSOL, which is a list of opinion bearing words in Spanish. The other one is the widely known opinion lexicon SentiWordNet. The result has been the filtered version of SentiWordNet by means the words that are in iSOL. The iSOL and SentiWordNet information that are in CRiSOL can be used jointly or indepently.

CRiSOL is composed by 8135 words of iSOL, from which 4434 are also linked with their polarity score in SentiWordNet.

How to cite:

Molina González, M. Dolores, Martínez Cámara, Eugenio, & Martín Valdivia, M. Teresa. (2015). CRiSOL: Opinion Knowledge-base for Spanish. Procesamiento Del Lenguaje Natural, 55, 143-150.
http://journal.sepln.org/sepln/ojs/ojs/index.php/pln/article/view/5226

Files of the resource:

crisol.tar.gz

emoti-sp

Resource type:

Lexicon

Description:

Linguistic resource for researching purposes in Sentiment Analysis on Spanish tweets. The lexicon is composed by 70 positive emoticons and 46 negative emoticons.

Files of the resource:

To download the resource you have to write an email to Salud M. Jiménez Zafra (sjzafra@ujaen.es) or Eugenio Martínez Cámara (emcamara@ujaen.es).

Hashtags-sp

Resource type:

Lexicon

Description:

Linguistic resource for researching purposes in Sentiment Analysis on Spanish tweets. The lexicon is composed by 172 positive Twitter hashtags and 127 negative Twitter hashtags.

Files of the resource:

To download the resource you have to write an email to Salud M. Jiménez Zafra (sjzafra@ujaen.es) or Eugenio Martínez Cámara (emcamara@ujaen.es).

eSOL

Resource type:

Lexicon

Description:

iSOL is a list of domain-dependent opinion signal words in Spanish. The domain is the set of words of movie reviews.

The elaboration of the list was performed using a corpus-based approach. In this case it selected the Spanish Movie Reviews corpus. The list is composed of 2,535 positive words and 5,639 negative words. For more information on how the list was developed see the paper: Semantic Orientation for Polarity Classification in Spanish Reviews (In revision).

Molina-González M.D., Martínez-Cámara, E., Martín-Valdivia, M. T. & Perea-Ortega, J. M. (2012). Semantic orientation for polarity classification in Spanish reviews. Expert Systems with Applications.
http://dx.doi.org/10.1016/j.eswa.2013.06.076

Resource files:

esol.tar.gz

iSOL

Resource type:

Lexicon

Description:

iSOL is a list of domain independent opinion signal words in Spanish.

For the elaboration of the resource it has begun from the list of words that the professors Bing Liu maintains (Bing Liu’s Opinion Lexicon). The word list has been automatically translated using the Reverso translator and subsequently corrected manually.

The list consists of 2,509 positive and 5,626 negative words. For more information on how the list was developed see the paper: Semantic Orientation for Polarity Classification in Spanish Reviews.

Reference

If you use iSOL, please, cite the following paper:

Molina-González, M. D., Martínez-Cámara, E., Martín-Valdivia, M. T., & Perea-Ortega, J. M. (2013). Semantic orientation for polarity classification in Spanish reviews. Expert Systems with Applications, 40(18), 7250-7257.

Files of the resource:

isol.tar.gz

SOL

Resource type:

Lexicon

Description:

SOL is a list of opinion signal words in Spanish independent of the domain.

For the elaboration of the resource it has begun with the list of words that maintains the professor Bing Liu (Bing Liu’s Opinion Lexicon). The list of words has been automatically translated using the translator Reverso.

The list consists of 1,397 positive and 3,151 negative words. For more information on how the list was developed see the article: Bilingual Experiments on an Opinion Comparable Corpus (in press).

Martínez-Cámara, E., Martín-Valdivia, M. T., Molina-Gonzalez, M. L. & Alfonso Ureña-López, L. (2013). Bilingual Experiments on an Opinion Comparable Corpus. Proceedings of the 4th Workshop on Computational Approaches to Subjectivity, Sentiment and Social Media Analysis
http://aclweb.org/anthology/W13-1612

Resource files:

sol.tar.gz

eSOLdomainGlobal

Resource type:

Lexicon

Description:

One of the main problems in Opinion Analysis is generating resources adapted for a specific domain. eSOLdomainGlobal is a set of lists of opinion signal words in Spanish that cover 8 different domains: cars, hotels, washing machines, books, mobile phones, music, computers and movies. The lists have been generated from the lexicon ISOL, and using a corpus-based approach taking the Spanish version of the SFU Review Corpus 8 lists have been generated.

Words

Positive

Negative

Cars

2528

5648

Hotels

2517

5636

Washers

2520

5639

Books

2529

5651

Mobile

2529

5657

Music

2538

5645

Computers

2527

5644

Films

2535

5648

Resource files:

eSOLdomainGlobal.rar