Background: Sentiment analysis (SA) is a subfield of natural language processing whose aim is to automatically classify the sentiment expressed in a free text. It has found practical applications across a wide range of societal contexts including marketing, economy, and politics. This review focuses specifically on applications related to health, which is defined as “a state of complete physical, mental, and social well-being and not merely the absence of disease or infirmity.” Objective: This study aimed to establish the state of the art in SA related to health and well-being by conducting a systematic review of the recent literature. To capture the perspective of those individuals whose health and well-being are affected, we focused specifically on spontaneously generated content and not necessarily that of health care professionals. Methods: Our methodology is based on the guidelines for performing systematic reviews. In January 2019, we used PubMed, a multifaceted interface, to perform a literature search against MEDLINE. We identified a total of 86 relevant studies and extracted data about the datasets analyzed, discourse topics, data creators, downstream applications, algorithms used, and their evaluation. Results: The majority of data were collected from social networking and Web-based retailing platforms. The primary purpose of online conversations is to exchange information and provide social support online. These communities tend to form around health conditions with high severity and chronicity rates. Different treatments and services discussed include medications, vaccination, surgery, orthodontic services, individual physicians, and health care services in general. We identified 5 roles with respect to health and well-being among the authors of the types of spontaneously generated narratives considered in this review: a sufferer, an addict, a patient, a carer, and a suicide victim. Out of 86 studies considered, only 4 reported the demographic characteristics. A wide range of methods were used to perform SA. Most common choices included support vector machines, naïve Bayesian learning, decision trees, logistic regression, and adaptive boosting. In contrast with general trends in SA research, only 1 study used deep learning. The performance lags behind the state of the art achieved in other domains when measured by F-score, which was found to be below 60% on average. In the context of SA, the domain of health and well-being was found to be resource poor: few domain-specific corpora and lexica are shared publicly for research purposes. Conclusions: SA results in the area of health and well-being lag behind those in other domains. It is yet unclear if this is because of the intrinsic differences between the domains and their respective sublanguages, the size of training datasets, the lack of domain-specific sentiment lexica, or the choice of algorithms.

Recently the Electric Network Frequency (ENF), one of the main traits of a power grid, had become increasingly popular in forensics since it is considered as a signature in multimedia recordings. By analyzing the ENF, it is possible to determine the time and location of a recording. In this paper, the ENF signals were classified using five different machine learning algorithms in order to detect the region of the origin of the ENF signals extracted from power and audio recordings coming from 10 different electric networks. Three sets of novel signal features are introduced and compared with the ones previously discussed in the literature. The improvement in the classification accuracy when a combination of the referent and novel feature sets was used ranges from 3% to 19% for the ENF signals extracted from power and audio recordings, respectively. Finally, the classifier with the highest achieved average accuracy was found to be Random Forest.

(1) Background: Aspect-based sentiment analysis (SA) is a natural language processing task, the aim of which is to classify the sentiment associated with a specific aspect of a written text. The performance of SA methods applied to texts related to health and well-being lags behind that of other domains. (2) Methods: In this study, we present an approach to aspect-based SA of drug reviews. Specifically, we analysed signs and symptoms, which were extracted automatically using the Unified Medical Language System. This information was then passed onto the BERT language model, which was extended by two layers to fine-tune the model for aspect-based SA. The interpretability of the model was analysed using an axiomatic attribution method. We performed a correlation analysis between the attribution scores and syntactic dependencies. (3) Results: Our fine-tuned model achieved accuracy of approximately $95\%$ on a well-balanced test set. It outperformed our previous approach, which used syntactic information to guide the operation of a neural network and achieved an accuracy of approximately $82\%$. (4) Conclusions: We demonstrated that a BERT-based model of SA overcomes the negative bias associated with health-related aspects and closes the performance gap against the state-of-the-art in other domains.

BACKGROUND Sentiment analysis (SA) is a subfield of natural language processing whose aim is to automatically classify the sentiment expressed in a free text. It has found practical applications across a wide range of societal contexts including marketing, economy, and politics. This review focuses specifically on applications related to health, which is defined as “a state of complete physical, mental, and social well-being and not merely the absence of disease or infirmity.” OBJECTIVE This study aimed to establish the state of the art in SA related to health and well-being by conducting a systematic review of the recent literature. To capture the perspective of those individuals whose health and well-being are affected, we focused specifically on spontaneously generated content and not necessarily that of health care professionals. METHODS Our methodology is based on the guidelines for performing systematic reviews. In January 2019, we used PubMed, a multifaceted interface, to perform a literature search against MEDLINE. We identified a total of 86 relevant studies and extracted data about the datasets analyzed, discourse topics, data creators, downstream applications, algorithms used, and their evaluation. RESULTS The majority of data were collected from social networking and Web-based retailing platforms. The primary purpose of online conversations is to exchange information and provide social support online. These communities tend to form around health conditions with high severity and chronicity rates. Different treatments and services discussed include medications, vaccination, surgery, orthodontic services, individual physicians, and health care services in general. We identified 5 roles with respect to health and well-being among the authors of the types of spontaneously generated narratives considered in this review: a sufferer, an addict, a patient, a carer, and a suicide victim. Out of 86 studies considered, only 4 reported the demographic characteristics. A wide range of methods were used to perform SA. Most common choices included support vector machines, naïve Bayesian learning, decision trees, logistic regression, and adaptive boosting. In contrast with general trends in SA research, only 1 study used deep learning. The performance lags behind the state of the art achieved in other domains when measured by F-score, which was found to be below 60% on average. In the context of SA, the domain of health and well-being was found to be resource poor: few domain-specific corpora and lexica are shared publicly for research purposes. CONCLUSIONS SA results in the area of health and well-being lag behind those in other domains. It is yet unclear if this is because of the intrinsic differences between the domains and their respective sublanguages, the size of training datasets, the lack of domain-specific sentiment lexica, or the choice of algorithms.

There has been a great interest in investigating relations between personality and language use on the web or social media. Most of the recent studies are based on mining the users’ information available online and then using machine learning algorithms to predict their personality characteristics. On the other hand, a few studies have relied on the traditional lexical hypothesis when exploring personality under the assumption that personality-related attributes could be obtained from dictionaries. However, little is known about personality structure from Twitter - do data strictly reflect personality structure as represented by personality models, or as unique personality semantic patterns. The aim of the study was to assess and interpret the personality adjective-based structure contained in tweets. The data were collected from an open-access „Tweet-sr“ Serbian Twitter linguistic corpus (Ljubešić & Klubička, 2014). Latent Dirichlet Allocation, a topic modeling technique, was conducted to extract topics and cosine similarity was used as a measure to determine topic similarities, as well as topic-personality dimensions’s similarities. The results showed that the optimal solution comprised four non-overlapping topics reflecting specific semantic structures. Topics did not replicate trait constructs but were modestly related to them. The largest similarities were found with Extraversion and Agreeableness, pointing out the conceptual importance of these traits when describing interpersonal behavior. Also, no inter-topic differences in category distributions were found, with the evaluation terms being the second most frequent in three topics. Although tweets are short-form text messages, they have the potential to communicate socially relevant information through personality descriptors.