Abstract:Purpose
Due to the large and fast growing sentiment analysis (SA) area recently, many new concepts and different nomenclatures have emerged without the desired organization. This confusion in the research field makes the understandability of the concepts hard and also hampers the comparison of different approaches. Thus, this paper aims to propose a hierarchical taxonomy to help the consolidation of SA area. The taxonomy aims at covering the addressed problems and methods in the SA field.
Design/methodology/… Show more
“…Categories of Sentiment Analysis (Jindal & Aron, 2021) The first method is to process text data based on various machine learning algorithms, which can be further broken down into supervised and unsupervised learning techniques (Jindal & Aron, 2021). Basically, machine learning methods extract keywords in text which would be regarded as features to learn and analyze (Rodrigues et al, 2018). The lexicon-based method assumes that synonyms have the same sentiment polarity and antonyms have opposite memory, and the text is analyzed by creating a universal dictionary (Birjali et al, 2021).…”
Section: Background Of Sentiment Analysismentioning
The objective of this project is to address the research question “What factor determines the success of an action movie?”, where the success factors are quantified as plot, music, plot, and director. Based on the research question, the author conducts a literature review of the development of the UK film industry, the impact of social media and film criticism on the film industry. To evaluate film reviews, sentiment analysis, which can efficiently convert text into structured data, is utilized in the project. In the methodology and the case study, the method of constructing the VADER model and the SVM model with the process including data processing, modelling, result correctness verification, and model evaluation are expounded and shown. Finally, the correctness of the two models is verified and the model output is visualized for analysis. The conclusion of the project is as follows: (1) Responding to the inquiry: In action films, the plot is the aspect that the audience pays the most attention to, but first the actors and music, and finally the director. (2) When compared to the experimental findings of Wang et al. (2020a), the project came to different conclusions, presumably as a result of the project’s limited access to movies and the change in the data processing approach. (3) Given that the two models employed in this research did well in the evaluation, they can be effectively applied to help production firms raise the caliber of action movies. (4) The models in the research can be easily developed and used to other domains because film review analysis can be compared with film reviews from other industries. However, some restrictions are also brought up, which offers suggestions for further work.
“…Categories of Sentiment Analysis (Jindal & Aron, 2021) The first method is to process text data based on various machine learning algorithms, which can be further broken down into supervised and unsupervised learning techniques (Jindal & Aron, 2021). Basically, machine learning methods extract keywords in text which would be regarded as features to learn and analyze (Rodrigues et al, 2018). The lexicon-based method assumes that synonyms have the same sentiment polarity and antonyms have opposite memory, and the text is analyzed by creating a universal dictionary (Birjali et al, 2021).…”
Section: Background Of Sentiment Analysismentioning
The objective of this project is to address the research question “What factor determines the success of an action movie?”, where the success factors are quantified as plot, music, plot, and director. Based on the research question, the author conducts a literature review of the development of the UK film industry, the impact of social media and film criticism on the film industry. To evaluate film reviews, sentiment analysis, which can efficiently convert text into structured data, is utilized in the project. In the methodology and the case study, the method of constructing the VADER model and the SVM model with the process including data processing, modelling, result correctness verification, and model evaluation are expounded and shown. Finally, the correctness of the two models is verified and the model output is visualized for analysis. The conclusion of the project is as follows: (1) Responding to the inquiry: In action films, the plot is the aspect that the audience pays the most attention to, but first the actors and music, and finally the director. (2) When compared to the experimental findings of Wang et al. (2020a), the project came to different conclusions, presumably as a result of the project’s limited access to movies and the change in the data processing approach. (3) Given that the two models employed in this research did well in the evaluation, they can be effectively applied to help production firms raise the caliber of action movies. (4) The models in the research can be easily developed and used to other domains because film review analysis can be compared with film reviews from other industries. However, some restrictions are also brought up, which offers suggestions for further work.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.