A First Approach to Mining Opinions as Multisets through Argumentation

Web 2.0 technologies have resulted in an exponential growth of text-based opinions coming from different sources (such as online news media, microblogging platforms, social networks, online review systems, etc.). The assessment of such opinions has gained considerable interest within several researc...

Descripción completa

Detalles Bibliográficos
Autores: Chesñevar, Carlos Ivan, González, María Paula, Grosse, Kathrin, Maguitman, Ana Gabriela
Tipo de recurso: artículo
Estado:Versión publicada
Fecha de publicación:2013
País:Argentina
Institución:Consejo Nacional de Investigaciones Científicas y Técnicas
Repositorio:CONICET Digital (CONICET)
Idioma:inglés
OAI Identifier:oai:ri.conicet.gov.ar:11336/12415
Acceso en línea:http://hdl.handle.net/11336/12415
Access Level:acceso abierto
Palabra clave:Argumentation
Opinion Mining
Egovernment
Datamining
https://purl.org/becyt/ford/1.2
https://purl.org/becyt/ford/1
Descripción
Sumario:Web 2.0 technologies have resulted in an exponential growth of text-based opinions coming from different sources (such as online news media, microblogging platforms, social networks, online review systems, etc.). The assessment of such opinions has gained considerable interest within several research communities in Computer Science, particularly in the context of modelling decision making processes. In this context, the scientific study of emotions in opinions associated with a given topic has become particularly relevant. Some approaches for assessing emotions in text-based opinions have been developed, resulting in promising software tools for sentiment analysis. In spite of the existence of such tools, assessing and contrasting text-based opinions is indeed a difficult task. On the one hand, complex opinions are built in many cases bottom up, emerging by aggregation from individual opinions posted online. On the other hand, contradictory and potentially inconsistent information might arise when contrasting such complex opinions. This article introduces an argument-based framework which allows to mine text-based opinions based on incrementally generated topics along with partially-ordered features, which provide a multidimensional comparison criterion. Given a topic, we will model an atomic opinion supporting it as a multiset (or bag) of terms. Atomic opinions can be aggregated, and related to alternative opinions, based on expanded topics. As a result, we will be able to obtain an “opinion analysis tree”, rooted in the first original topic.