The Sentiment Hidden in Italian Texts Through the Lens of A New Dictionary
[EN] The aim of this work is to propose a strategy to classify texts (or parts of them) in an ordinal emotional scale to gauge a sentiment indicator in every domain. In particular, we develop a new dictionary for the Italian language which is built using an objective method where the polarities of s...
| Autores: | , , , , |
|---|---|
| Tipo de recurso: | capítulo de libro |
| Fecha de publicación: | 2018 |
| País: | España |
| Institución: | Universitat Politècnica de València (UPV) |
| Repositorio: | RiuNet. Repositorio Institucional de la Universitat Politécnica de Valéncia |
| Idioma: | inglés |
| OAI Identifier: | oai:riunet.upv.es:10251/111926 |
| Acceso en línea: | https://riunet.upv.es/handle/10251/111926 |
| Access Level: | acceso abierto |
| Palabra clave: | Web data Internet data Big data QCA PLS SEM Conference Text analysis Sentiment analysis |
| Sumario: | [EN] The aim of this work is to propose a strategy to classify texts (or parts of them) in an ordinal emotional scale to gauge a sentiment indicator in every domain. In particular, we develop a new dictionary for the Italian language which is built using an objective method where the polarities of synonyms and antonyms are accounted for in an iterative process. To build our sentiment indicator negations and intensifiers are also used, thus considering the context in which the single word is written. We apply our new dictionary to extract the sentiment from a set of around 40 issues of the Bank of Italy quarterly Economic Bulletin. Our results show that our strategy is able to correctly identify the sentiment expressed in the Bulletins, which is correlated to the main macroeconomic variables (such as national GDP, investment, consumption or unemployment rate). Our analysis shows that sentiment represents not only an evaluation of the stylistic way in which texts are written, but also a valid synthesis of all the external factors analysed in the same document. |
|---|