Skip to content

Latest commit

 

History

History
executable file
·
28 lines (21 loc) · 3.75 KB

README.md

File metadata and controls

executable file
·
28 lines (21 loc) · 3.75 KB

Latin_Sentiment_Lexicons

This repository contains a new set of sentiment lexicons of Latin adjectives and nouns which content is summarized in the table below.

  • cross-lingual-projection.tsv was generated using a cross-language projection method starting from the English lexicon by Cho et al. (2014).
  • induction-kNN.tsv was induced from distributed vector representations of words with a k-NN algorithm implementation (Sidarenka, 2019).
  • gold_3classes.tsv and gold_5classes.tsv were created by two experts of Latin language and culture following a multi-stage process and an extensive reconciliation phase. In the first each entry has 1 out of 3 possible score: positive, negative, neutral. The second has a five-way classification: 1 (fully positive), 0.5 (somewhat positive), 0 (neutral), -0.5 (somewhat negative), -1 (fully negative).
  • silver.tsv was built by deriving new lexicon entries through synonym, antonym and derivational relations with the entries in the gold standard. Synonyms and antonyms were taken from the Latin dictionary compiled by Skřivan (1890). Derivational relations were extracted from the database of Word Formation Latin. Graphical variants of lemmas present in the gold standard were added to the silver standard as well.
  • LatinAffectus.tsv was created by merging the gold standard and the silver standard. It has been linked to the LiLa:Linking Latin Knowledge Base.
  • Extension.tsv: new set of lemmas associated with a polarity score. It is made of a new Gold Standard and a new Silver Standard.
  • LatinAffectusv2.tsv, LatinAffectusv3.tsv, LatinAffectusv4.tsv: extensions of LatinAffectus, the most up-to-date version is v4.

How to cite

Papers

Sprugnoli, R., Passarotti, M., Corbetta, D., & Peverelli, A. (2020). Creating, Evaluating and Extending Sentiment Lexicons for Latin. In Proceedings of the Twelfth International Conference on Language Resources and Evaluation (pp. 3071-3079). European Language Resources Association (ELRA). ([PDF])(https://www.aclweb.org/anthology/2020.lrec-1.376.pdf)

For LatinAffectus, its modeling and linking: Sprugnoli, R., Mambrini, F., Moretti, G., Passarotti, M. (2020). Towards the Modeling of Polarity in a Latin Knowledge Base. In Proceedings of the 3rd Workshop on Humanities in the Semantic Web (WHiSe III). ([PDF])(http://ceur-ws.org/Vol-2695/paper7.pdf)

Dataset

Rachele Sprugnoli, Giovanni Moretti, Marco Passarotti, Daniela Corbetta, Andrea Peverelli. (2020). CIRCSE/Latin_Sentiment_Lexicons: First release (Version v1.0.0) [Data set]. Zenodo. http://doi.org/10.5281/zenodo.3865154

LatinAffectus is also available in the ILC4CLARIN repository of CLARIN-IT: http://hdl.handle.net/20.500.11752/OPEN-527.

Copyright

Creative Commons Licence
These resources are licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.

Funding

The LiLa: Linking Latin project has received funding from the European Research Council (ERC) under the European Union’s Horizon 2020 research and innovation programme – Grant Agreement No. 769994.