ABSTRACT
Despite the popularity and efficiency of dictionary-based sentiment analysis (DSA) for public health research, limited empirical evidence has been produced about the validity of DSA and potential harms to the validity of DSA. A random sample of a second-hand Ebola tweet dataset was used to evaluate the validity of DSA compared to the manual coding approach and examine the influences of textual features on the validity of DSA. The results revealed substantial inconsistency between DSA and the manual coding approach. The presence of certain textual features such as negation can partially account for the inconsistency between DSA and manual coding. The findings imply that scholars should be careful and critical about findings in disease-related public health research that use DSA. Certain textual features should be more carefully addressed in DSA.
Subject(s)
Communicable Diseases, Emerging , Social Media , Attitude , Humans , Public Health , Sentiment AnalysisABSTRACT
The Covid pandemic has become a serious public health challenge for people across India and other nations. Nowadays, people rely on the online reviews being shared on different review sites to gather information about hospitals like the availability of beds, availability of ventilators, etc. However, since these reviews are large in number and are unstructured, patients struggle to get accurate and reliable information about the hospitals, due to which they end up taking admission into a hospital which might not be appropriate for the specific treatment they require. This paper employs the use of sentiment analysis to understand various online reviews of hospitals and provide valuable information to the patients. Approximately 30,000 + reviews were collected from more than 500 hospitals. The broad objective of the study is to give the patients a comprehensive and descriptive rating of the hospitals based on the online reviews given by different patients. In addition to providing a comprehensive summary, the study has conducted aspect-based analysis where it compares the hospitals based on four different aspects of the hospital viz. "Doctors' services", "Staff's services", "Hospital facilities", and "Affordability". The database containing aspect-based ratings of the hospitals will be of great value to the patients by allowing them to compare and select the best hospital based on the optimum fit of the aspects of their preference.
ABSTRACT
Describimos la aplicación de la tecnología de procesamiento de lenguaje natural (NLP) al análisis del lenguaje subjetivo. En particular, nos concentramos en la problemática de la clasificación de opinión de material textual extraído de fuentes de datos relacionados con negocios. Estudiamos la derivación de los valores de opiniones de palabras a partir del recurso léxico SentiWordNet y utilizamos estos valores para la interpretación de texto con el objetivo de obtener la valoración de una opinión a partir de sus palabras y frases. Utilizamos características de las palabras para inducir un clasificador basado en el uso de Máquinas de Vectores de Soporte que alcanzan resultados acordes con el estado del arte. También mostramos experimentos preliminares en los que el uso de resúmenes de opiniones ofrece ventaja competitiva para el problema de clasificación respecto del uso de documentos completos cuando los documentos son extensos y contienen material tanto subjetivo como no-subjetivo.
We describe the application of natural language processing (NLP) technology to the analysis of subjective language. In particular we concentrate on the problem of opinion classification of textual material extracted from business-related data-sources. We study the derivation of sentiment values for words from the SentiWordNet lexicalresource and use them for text interpretation to produce word, sentence, and text based sentiment features for opinion classification. We use word-based and sentiment basedfeatures to induce a classifier based on the use of Support Vector Machinesachieving state of the art results. We also show preliminary experiments where the use of summaries before opinion classification provides competitive advantage over the use of full documents when the documents are long and contain both subjective andnon-subjective material.
Subject(s)
Language , Natural Language Processing , Software , PsychologyABSTRACT
Describimos la aplicación de la tecnología de procesamiento de lenguaje natural (NLP) al análisis del lenguaje subjetivo. En particular, nos concentramos en la problemática de la clasificación de opinión de material textual extraído de fuentes de datos relacionados con negocios. Estudiamos la derivación de los valores de opiniones de palabras a partir del recurso léxico SentiWordNet y utilizamos estos valores para la interpretación de texto con el objetivo de obtener la valoración de una opinión a partir de sus palabras y frases. Utilizamos características de las palabras para inducir un clasificador basado en el uso de Máquinas de Vectores de Soporte que alcanzan resultados acordes con el estado del arte. También mostramos experimentos preliminares en los que el uso de resúmenes de opiniones ofrece ventaja competitiva para el problema de clasificación respecto del uso de documentos completos cuando los documentos son extensos y contienen material tanto subjetivo como no-subjetivo.(AU)
We describe the application of natural language processing (NLP) technology to the analysis of subjective language. In particular we concentrate on the problem of opinion classification of textual material extracted from business-related data-sources. We study the derivation of sentiment values for words from the SentiWordNet lexicalresource and use them for text interpretation to produce word, sentence, and text based sentiment features for opinion classification. We use word-based and sentiment basedfeatures to induce a classifier based on the use of Support Vector Machinesachieving state of the art results. We also show preliminary experiments where the use of summaries before opinion classification provides competitive advantage over the use of full documents when the documents are long and contain both subjective andnon-subjective material.(AU)
