Semi-Supervised Generative Adversarial Network for Sentiment Analysis of drug reviews

<div>Sentiment analysis has become a very popular research topic and covers a wide range of domains such as economy, politics and health. In the pharmaceutical field, automated analysis of online user reviews provides information on the effectiveness and potential side effects of drugs, which could be used to improve pharmacovigilance systems. Deep learning approaches have revolutionized the field of Natural Language Processing (NLP), achieving state-of-the-art results in many tasks, such as sentiment analysis.</div><div>These methods require large annotated datasets to train their models. However, in most real-world scenarios, obtaining high-quality labeled datasets is an expensive and time-consuming task. In contrast, unlabeled texts task can be, generally, easily obtained. </div><div>In this work, we propose a semi-supervised approach based on a Semi-Supervised Generative Adversarial Network (SSGAN) to address the lack of labeled data for the sentiment analysis of drug reviews, and improve the results provided by supervised approaches in this task.</div><div>To evaluate the real contribution of this approach, we present a benchmark comparison between our semi-supervised approach and a supervised approach, which uses a similar architecture but without the generative adversal setting. </div><div>Experimental results show better performance of the semi-supervised approach when annotated reviews are less than ten percent of the training set, obtaining a significant improvement for the classification of neutral reviews, the class with least examples. To the best of our knowledge, this is the first study that applies a SSGAN to the sentiment classification of drug reviews. Our semi-supervised approach provides promising results for dealing with the shortage of annotated dataset, but there is still much room to improvement.</div>

Semi-Supervised Generative Adversarial Network for Sentiment Analysis of drug reviews

<div>Sentiment analysis has become a very popular research topic and covers a wide range of domains such as economy, politics and health. In the pharmaceutical field, automated analysis of online user reviews provides information on the effectiveness and potential side effects of drugs, which could be used to improve pharmacovigilance systems. Deep learning approaches have revolutionized the field of Natural Language Processing (NLP), achieving state-of-the-art results in many tasks, such as sentiment analysis.</div><div>These methods require large annotated datasets to train their models. However, in most real-world scenarios, obtaining high-quality labeled datasets is an expensive and time-consuming task. In contrast, unlabeled texts task can be, generally, easily obtained. </div><div>In this work, we propose a semi-supervised approach based on a Semi-Supervised Generative Adversarial Network (SSGAN) to address the lack of labeled data for the sentiment analysis of drug reviews, and improve the results provided by supervised approaches in this task.</div><div>To evaluate the real contribution of this approach, we present a benchmark comparison between our semi-supervised approach and a supervised approach, which uses a similar architecture but without the generative adversal setting. </div><div>Experimental results show better performance of the semi-supervised approach when annotated reviews are less than ten percent of the training set, obtaining a significant improvement for the classification of neutral reviews, the class with least examples. To the best of our knowledge, this is the first study that applies a SSGAN to the sentiment classification of drug reviews. Our semi-supervised approach provides promising results for dealing with the shortage of annotated dataset, but there is still much room to improvement.</div>

Improving RE-SWOT Analysis with Sentiment Classification: A Case Study of Travel Agencies

Nowadays, many companies collect online user reviews to determine how users evaluate their products. Dalpiaz and Parente proposed the RE-SWOT method to automatically generate a SWOT matrix based on online user reviews. The SWOT matrix is an important basis for a company to perform competitive analysis; therefore, RE-SWOT is a very helpful tool for organizations. Dalpiaz and Parente calculated feature performance scores based on user reviews and ratings to generate the SWOT matrix. However, the authors did not propose a solution for situations when user ratings are not available. Unfortunately, it is not uncommon for forums to only have user reviews but no user ratings. In this paper, sentiment analysis is used to deal with the situation where user ratings are not available. We also use KKday, a start-up online travel agency in Taiwan as an example to demonstrate how to use the proposed method to build a SWOT matrix.

Dynamic Restaurants Quality Mapping Using Online User Reviews

Millions of users post comments to TripAdvisor daily, together with a numeric evaluation of their experience using a rating scale of between 1 and 5 stars. At the same time, inspectors dispatched by national and local authorities visit restaurant premises regularly to audit hygiene standards, safe food practices, and overall cleanliness. The purpose of our study is to analyze the use of online-generated reviews (OGRs) as a tool to complement official restaurant inspection procedures. Our case study-based approach, with the help of a Python-based scraping library, consists of collecting OGR data from TripAdvisor and comparing them to extant restaurants’ health inspection reports. Our findings reveal that a correlation does exist between OGRs and national health system scorings. In other words, OGRs were found to provide valid indicators of restaurant quality based on inspection ratings and can thus contribute to the prevention of foodborne illness among citizens in real time. The originality of the paper resides in the use of big data and social network data as a an easily accessible, zero-cost, and complementary tool in disease prevention systems. Incorporated in restaurant management dashboards, it will aid in determining what action plans are necessary to improve quality and customer experience on the premises.

Analyzing the Voice of Customer through online user reviews using LDA: Case of Moroccan mobile banking applications

Nowadays, most of the organizations make their mobile applications available through different stores, such as Google Play Store, Apple App Store, and Windows Phone Store. Banks and financial institutions have also provided mobile applications for their customers. These app stores not only allow applications to be downloaded, but they also permit users to leave comments and reviews. In this paper, we will start first by looking at eight Moroccan mobile banking applications in the Google Play Store. Data that hasn’t been exploited by Moroccan banks yet. Once the preprocessing phase is complete, we will examine and analyze user reviews using Latent Dirichlet allocation (LDA) to extract and identify topics. Topics discovered focus mainly on Security, services, quality, and interface. While customer reviews can influence future demand, they can also be used by managers to improve their services and customer experience.