According to the Pan American Health Organization (PAHO) [1], syphilis is a Sexually Transmitted Infection (STI) caused by the bacterium Treponema pallidum. Left untreated or inadequately treated, the disease can lead to severe health problems and even cause death in infected persons. In addition, babies born to untreated mothers are likely to acquire congenital syphilis during pregnancy (mother-to-child transmission, MTCT) and childbirth [2].

Syphilis diagnosis usually involves serological testing, and the disease can be easily treated with antibiotics (i.e., penicillin) since it is a bacterial infection [2]. In addition, the correct use of condoms is an effective and simple method to prevent syphilis from being spread through sexual intercourse [1].

Nearly 6.3 million cases of syphilis were reported among people aged 15–49 years globally in 2016 [1]. According to World Health Organization (WHO) [1], the disease disproportionately affects key populations. For instance, the incidence of syphilis is as high as 27% among men who have sex with men (MSM) and 14% among female sex workers. However, across the Americas, key populations have not been screened for syphilis in 35% of countries, even though serology tests for STIs like HIV are performed. This reinforces the fact that syphilis remains a neglected disease [1]. Additionally, WHO estimates 930,000 yearly cases of syphilis in pregnant women with active infection (transmissible during pregnancy), resulting in approximately 350,000 adverse birth outcomes annually, i.e., malformations and preterm births.

In Brazil, the 2021 Epidemiological Bulletin released yearly by the Brazilian Ministry of Health (MoH) reported that the detection rate of acquired syphilis had risen considerably until 2018, when it peaked at 76.4 cases per 100,000 population [3]. In contrast, the acquired syphilis rate dropped to 54.5 cases per 100,000 population in 2020. Of note, the infection has been listed as a notifiable disease since 2010. In 2018, the rate of congenital syphilis rate reached 9.0 cases per 1000 live births, with a drop in the following years, reaching 7.7 cases per 1000 live births in 2020 and with 186 children under one-year-old dying from syphilis, which corresponds to a mortality coefficient of 6.5 deaths per 100,000 live births in 2020. Meanwhile, the detection rate of syphilis in pregnant women reached 21.8 cases per 1000 live births in 2019 and decreased to 21.6 per 1000 live births in 2020.

In this context, public awareness campaigns play a pivotal role in disseminating relevant information on health promotion and disease prevention, especially to key populations ^[4][5][4,5]. Campaigns centered on themes around syphilis (i.e., transmission, prevention, symptomatology, disease progression, diagnosis, and treatment) may maximize the chance the disease is caught at early stages, thus allowing prompt and simplified treatment, and minimizing a person’s risk of becoming infected and spreading the infection.

As part of the fight against the rising incidence of syphilis in Brazil, the MoH launched the “Syphilis No!” Project (SNP). One of the main intervention actions of this project throughout the national territory was developing a massive and universal public health communication campaign throughout the country, in addition to specific actions for priority population groups.

Through the “Syphilis No!” Project the syphilis epidemic has been tackled through two strategic lines: (1) reinforcing universal actions of the National Brazilian Health System (SUS) and (2) implementing specific ones to 100 municipalities chosen by the Ministry of Health as priorities for the response to congenital syphilis, as in 2015, they represented 68.95 of the number of congenital syphilis cases in Brazil [6].

The universal line of intervention included the acquisition and distribution of supplies for testing and treatment (crystalline and benzathine penicillin), enhancing the STI laboratories network and situation rooms for epidemiological surveillance, educommunication strategies [7], social interventions, and awareness campaigns performed to face syphilis in that period [6].

The most important communication intervention of the SNP occurred between 2018 and 2019, namely the “Syphilis No!” campaign, which emphasized the concept “Test, Treat and Cure” to alert the population about the availability of the rapid syphilis test or VDRL test at any Primary Health Care (PHC) unit of the SUS.

During this period, the organizers produced and disseminated a large amount of material through television, radio, streaming platforms, printed media, magazines, posters, informative booklets, and stickers. Internet sites, specifically directed toward pregnant women, disseminated related content and other content strategically emerged within news coverage and on social networks, relationship apps, and digital pages of magazines. In addition, digital influencers made sponsored posts on their social networks [8].

Other studies have shown that the intervention actions of the SNP had positive effects, as the drop in the number of reported syphilis cases in the country and a change in the trend was observed after 2018 [6], when the Project started. Andrade et al. [9] demonstrate the hypothesis that the SNP has influenced the decline in hospitalizations for congenital syphilis in Brazilian municipalities as of May 2018. De Morais Pinto and colleagues [8] found that the SNP influenced the increase of testing nationwide, which consequently expanded diagnosis and treatment, which positively impacted the reduction of syphilis cases in the country.

In recent years, a vast body of research has been intensively exploring sentiment analysis applied to online news. Below, thwe researchers hhighlighted some studies found in the literature.

Lei and colleagues ^[10][16] proposed a system for sentiment detection in online news items. This system is based on document selection, part-of-speech (POS) annotations, and sentiment lexicon generation. The presented approach showed better results than a Support Vector Machine (SVM) based classifier.

Li et al. ^[11][17] introduced a Weighted Multi-label Classification Model (WMCM) for multi-label classification that applied weightings to the files used in the training process. In addition, another method was also used for checking different sentiments associated with an exact word at a semantic level.

In Bai’s work ^[12][18], sentiment analysis regarding online news was used to infer customer sentiment. The author proposes a hybrid heuristic using the Tabu search metaheuristic combined with the probabilistic Markov Envelope model to extract the dependencies between words and assemble a sentiment-based dictionary.

Rao et al. ^[13][19] proposed algorithms to automatically construct an emotion-based dictionary and a method to generate a topic-based dictionary. The method developed has some pertinent features, as it is language-independent and allows granulation, scale, and data volume adjustment.