INTRODUCTION
Individuals suffering from enuresis often struggle to discuss their condition openly due to significant stigma and embarrassment. This reluctance is exacerbated by the perception that enuresis is socially taboo and shameful issue. For instance, 60.6% of Austrian respondents considered urinary incontinence, a symptom of enuresis, more embarrassing than conditions such as depression or cancer [1,2]. This suggests that those with enuresis may experience heightened levels of shame and hesitancy in seeking help and thus, it leads to underreporting and untreated symptoms. Moreover, enuresis is often kept secret not only by the patient but also their family, which can result in substantial psychological stress for both the child and their family [3].
In addition, primary nocturnal enuresis (PNE), defined as experiencing nocturnal enuresis (NE) for more than 6 months, presents additional challenges. Research indicates that 34.5% of college students with PNE were scolded and punished, with 80% not receiving active treatment despite the fact that such punitive measures are ineffective and can lead to psychological issues [4]. This reluctance to discuss enuresis openly can hinder effective diagnosis and management, and highlights the need for healthcare professionals to adopt a sensitive and proactive approach to this condition.
Given the negative social perceptions surrounding enuresis, Text-mining techniques, a method for extracting useful knowledge by uncovering hidden structures or relationships within natural language data [5], can be an effective method for identifying and understanding public perceptions of enuresis. Online communities, in particular, provide anonymity and allow participants to share their experiences more openly without fear of stigma or judgment. This method has proven particularly advantageous for marginalized groups such as LGBTQ individuals and patients with rare diseases [6]. This is especially useful for exploring perceptions of conditions that patients and their families may not feel comfortable discussing offline.
Text-mining data from online communities offers several benefits. First, it allows for the collection of raw, honest expressions from individuals who may be unaware that they are participating in research. Second, it takes less time and money to collect and analyze data than traditional methods. Third, it enables the gathering of diverse samples from geographically dispersed participants so that it provides insights into broader cultural context. Despite these advantages, the use of online community data for studying disease perceptions remains quite limited. Some studies have analyzed online post and search activity to understand the impact of global campaigns and social media on public awareness of disease like coronavirus disease 2019 [7], inflammatory bowel disease [8], and sepsis [9].
In this regard, this study collects data from online communities in South Korea and the United States (US) to investigate current awareness and perceptions of enuresis.
For this purpose, this study employs text-mining techniques including Latent Dirichlet Allocation (LDA) topic modeling and visualization using pyLDAvis to extract hidden topics and identify key issues. By analyzing data from these 2 cultural contexts, the purpose of this study is to compare and contrast the differences in perceptions of enuresis across South Korea and the US.
MATERIALS AND METHODS
Text-Mining
Text-mining is the process of converting large volumes of unstructured text into structured data so that it can serve as a valuable source of information. This process helps to discern the meaning of the data by analyzing the frequency of keyword usage [10]. One advanced method of text-mining is topic modeling, which infers associations between words or sentences within the text data of a document and identifies significant common topics as probability distributions. Among various topic modeling techniques, LDA is the most widely used when employing a probabilistic approach to uncover hidden topics within the text [11]. The entire research process is summarized in (Fig. 1).
Data Collection and Text Preprocessing
In this study, text data were collected from online community posts to investigate public perception of enuresis. For data collection, this study focused on the content, titles, URLs, and names of the subcommunities where posts were uploaded on Naver Cafe in South Korea and Reddit in the US. These platforms were chosen for big data analysis across various fields such as substance use [12], suicide attempts [13], and financial imagery [14] due to their large, diverse user bases, extensive subcommunities, high engagement level and the anonymity they provide, which fosters open and candid discussions.
First, both Naver Cafe and Reddit have large user bases. As of February 2024, Naver Cafe had 2.38 million active users [15], while Reddit had 42 million logged-in users in Q2 2024 [16]. The second reason for selecting these platforms is their shared features of anonymity and support for subcommunities. Reddit, in particular, offers a unique environment where large groups can discuss health conditions freely and anonymously since topics are often difficult to survey due to cultural stigmas, especially in the health sector [17]. The subcommunities on both Naver Cafe and Reddit allow users to identify specific spaces where enuresis-related topics are discussed. On Naver Cafe, users post in ‘café,’ which are subcommunities created around specific interests or purposes. Similarly, Reddit users post in subcommunities called ‘Subreddit’ which are categorized by various factors such as interests. By examining the names of the cafes on Naver Cafe and the subreddits on Reddit, this study aims to identify the specific online spaces where discussions about enuresis took place.
Data collection was performed using the APIs provided by Naver Cafe developers [18] and Reddit, respectively. A Client ID for both platforms were obtained and searched for the term ‘nocturnal enuresis’ on Naver Cafe and the terms ‘enuresis’ and ‘nocturia’ on Reddit. The choice of keywords differed between the platforms due to variations in language usage. In South Korea, the term enuresis, which refers to the loss of discrete amounts of urine during sleep in children older than 5 years [19], the term nocturia, which refers to waking one or more times at night to void urine [20], and the term nocturnal enuresis, which refers to involuntary urination during sleep [21], are commonly used interchangeably under the umbrella term nocturnal enuresis (Ya-nyo-jeung). On Reddit, ‘enuresis’ and ‘nocturia’ were the most commonly used terms related to the condition.
A total of 1,000 posts were collected from Naver Cafe and 464 posts from Reddit. Since users on both platforms often composed their titles and content as a single cohesive post, this study used the pandas library to combine them for analysis. Preprocessing of the combined data included removing special characters and URLs, cleaning whitespace and duplicate characters, conducing morphological analysis and base type cleaning, as well as word removal and tokenization.
Word Cloud and LDA Topic Modeling Visualization
Word cloud
This study first visualized the tokenized data incorporating post contents and titles in the form of a word cloud. A word cloud is a graphical tool that provides a snapshot of the thematic focus of a field by visually highlighting the most common terms [22]. The resulting word cloud shows at a glance which words were most commonly used by users in each platform when posting about enuresis.
LDA topic modeling and visualization
This study created a dictionary and corpus and performed 15 rounds of LDA training. Initially, there was significant overlap in words across topics. To address this, this study employed the Gensim library to automate the optimization of alpha and beta parameters of the LDA model training. The training was experimented with varying the number of topics from 3 to 10 and the number of included words from 5 to 30 in increments of 5. Ultimately, 5 topics yielded the most even distribution on the Intertopic Distance Map, while 30 words provided the most context with minimal word duplication. Thus, visualization was conducted using the pyLDAvis model and Genism function based on 5 topics and 30 words per topic.
Analysis of Communities Names
Categorization and frequency visualization
To understand the types of communities where individuals were posting about enuresis, this study conducted a frequency analysis and visualization using the names of communities in each online platform. This analysis aimed to identify which spaces or groups users felt comfortable discussing enuresis-related questions and opinions, as well as the topics they considered relevant to the enuresis issues.
After manually categorizing the terms used in community names, the pandas library and the matplotlib.pylot module were employed to calculate the frequency of each category and visualize the results in a bar graph. Since community names in Naver Cafe often contain numerous proper nouns and special characters, this study preprocessed them for clarity. In contrast, while community names of Reddit generally contain fewer words, they sometimes feature titles that obscure the community’s focus. To address this, this study visited the communities directly through the collected URLs and categorized them accordingly.
RESULTS
Analysis of Post data in Naver Cafe and Reddit
Word cloud visualization
The tokenized data, which incorporates post content and titles, was visualized in a word cloud format as shown in Fig. 2.
The Naver Cafe word cloud highlights a range of terms related to basic enuresis symptoms including ‘urine,’ ‘bedwetting,’ ‘night,’ ‘toilet training,’ ‘symptom,’ ‘sleep,’ ‘mistake,’ ‘adult,’ and ‘child’. It also feature words associated with treatment facilities such as ‘hospital,’ ‘oriental medicine clinic,’ and ‘treatment’ as well as terms that suggest recommendations from medial institutions or treatment including ‘recommend,’ ‘try,’ ‘seeing a doctor,’ and ‘please.’
In contrast, the Reddit word cloud includes terms describing symptoms, such as ‘sleep,’ ‘night,’ ‘child,’ ‘im,’ ‘bedwetting,’ ‘nocturia,’ and ‘symptom.’ However, it stands out by incorporating more contextualized words, such as ‘day,’ ‘year,’ as well as references to body parts, such as ‘bladder’ and ‘pelvic.
A comparative analysis of the word clouds indicates that Naver Cafe serves primarily as an information exchange platform for individuals seeking solutions to enuresis, while Reddit functions more as a communication space for sharing daily experiences and emotions throughout the treatment process.
Results of LDA topic modeling and visualization
Figs. 3 and 4 present the results of the topic modeling and visualization based on λ=0.4. This value was selected as it best illustrates the best topic characteristics while minimizing meaningless redundant values. The included words are listed in order of frequency.
In the analysis of topic 1 titled Reporting adults enuresis symptoms from the Naver Cafe data, keywords such as ‘adult,’ ‘night,’ ‘frequent urination,’ ‘cause,’ ‘treatment,’ ‘oriental medicine,’ ‘sunday,’ ‘frequent,’ ‘urine,’ and ‘symptoms’ were prominent. These terms suggest that a significant interest in the symptoms and descriptions of adult enuresis, with a particular focus on herbal treatment methods.
Topic 2 featured numerous references to weather condition and place names in South Korea, such as ‘weather,’ ‘Jamsil,’ ‘Nowon,’ ‘Hyundai,’ ‘system,’ ‘stay,’ ‘Namyangju,’ ‘Busan,’ ‘spend,’ ‘day,’ ‘winter,’ ‘cool,’ and ‘cold.’ This pattern reflects the sharing of daily life experiences alongside discussions of illness and symptoms, which is characteristic of community-based platforms like Naver Cafe. In addition, the recurrence of keywords related to oriental clinic indicates sustained interest in traditional treatment.
Topic 3 was titled with requesting treatment referrals and sharing experiences since keywords including ‘hospital,’ ‘recommendation,’ ‘please,’ ‘test,’ ‘treatment,’ ‘examination,’ ‘go to see’ appeared alongside ‘enuresis.’ This suggests that the posts were primarily focused on seeking recommendations for urologists, hospitals, pediatricians, and oriental medicine clinics for enuresis treatment, with a notable desire to share personal experiences with these treatments.
Topic 4 includes ‘diaper,’ ‘dawn,’ ‘mistake,’ ‘sleep,’ ‘take off,’ ‘cognition,’ ‘toilet,’ ‘husband,’ and ‘cystitis,’ which indicate descriptions and concerns about nocturnal enuresis symptoms, sometimes on behalf of family members. Thus, it was titled with providing detailed symptom descriptions and family concerns.
Topic 5 contains keywords such as ‘toilet training,’ ‘symptoms,’ ‘urine,’ ‘child,’ ‘disease,’ ‘improvement,’ ‘bedwetting,’ ‘suspicion,’ ‘week,’ ‘age,’ ‘control,’ and ‘months’. These terms suggest detailed descriptions of pediatric enuresis symptoms, growth stages, and requests for diagnoses from others, contrasting with the adult focus of topic 4, which is titled with seeking pediatric enuresis diagnoses.
For the Reddit title and post data shown in Fig. 4, topic 1 contains includes keywords such as ‘day,’ ‘im,’ ‘time,’ ‘sleep,’ ‘get,’ ‘like,’ ‘year,’ ‘nocturia,’ ‘night,’ ‘feel,’ ‘would,’ ‘week,’ ‘started,’ ‘hour,’ ‘pee,’ ‘since,’ and ‘month.’ These keywords reflect how individuals describe their personal experiences with enuresis with the focus on the daily and temporal aspects of their symptoms.
Topic 2 is characterized by a significant number of terms related to crime and traumatic events including ‘mother,’ ‘killer,’ ‘serial,’ ‘murder,’ ‘abuse,’ ‘victim,’ ‘girl,’ ‘feeding,’ ‘crime,’ ‘child,’ ‘infant,’ ‘abused,’ ‘maternal,’ ‘killed,’ ‘killing,’ ‘sexual,’ ‘compared,’ ‘sexually,’ ‘behavior,’ and ‘homicide.’ This suggests infer that discussions of enuresis symptoms may be triggered by anxiety or stress related to crime and traumatic events.
Topic 3 is more medically focused with precise terms related to body parts or disease, such as ‘pelvic,’ ‘prostate,’ ‘symptom,’ ‘pain,’ ‘floor,’ ‘urinary,’ ‘treatment,’ ‘kidney,’ ‘prostatitis,’ ‘dysfunction,’ ‘disease,’ ‘link,’ ‘surgery,’ ‘urologist,’ ‘antibiotic,’ ‘physical,’ ‘medicine,’ ‘urination,’ ‘discomfort,’ ‘including,’ ‘muscle,’ ‘urology,’ ‘ejaculation,’ ‘testicle,’ and ‘testicle.’ Thus, this topic was titled with description of enuresis treatment experiences and familiarity with the disease. These terms suggest that the posts are written by individuals with a deep understanding of the condition, possibly those who have been diagnosed or have lived with enuresis for an extended period compared to topic 1.
Topic 4 includes the keywords such as ‘bedwetting,’ ‘alarm,’ ‘insulin,’ ‘fat,’ ‘child,’ ‘protein,’ ‘food,’ ‘sugar,’ ‘resistance,’ ‘eating,’ ‘carbs,’ ‘level,’ ‘increase,’ ‘mattress,’ ‘bedtime,’ ‘reduce,’ ‘system,’ ‘body,’ ‘carbohydrate,’ ‘diet,’ ‘fasting,’ ‘control,’ ‘intake,’ ‘meal,’ ‘quality,’ ‘amount,’ etc. The repeated mention of ‘child’ suggests that this topic addresses lifestyle management not only for adults but also for children, with a focus on diet and related factors.
Topic 5 contains many place names and other words that have been preprocessed and thus are difficult to interpret. However, it also includes terms like ‘antihistamine,’ ‘amino,’ ‘visceral,’ ‘vascular,’ ‘bcaa,’ ‘hub,’ and ‘resveratrol’, indicating an interest in medicine and treatment methods.
After analyzing and comparing topic modeling, 3 significant findings can be discussed.
Online community as a psychological outlet: In both community channels, topic 1 focuses on discussing enuresis symptoms, suggesting that online communities serve as an important psychological outlet for sharing personal experiences and daily life related to enuresis in both South Korea and the US.
Differences Between Naver Cafe and Reddit: A stark difference is observed between the 2 platforms when comparing topics 3, 4, and 5. On Naver Cafe, discussions are dominated by concerns from family members including parents worried about their children’s enuresis, requests for diagnoses, and recommendations for treatment centers. This indicates a tendency among Korean users to seek information or shared experiences before consulting a healthcare professional. In contrast, Reddit users in the US focus on medical information regarding the treatment process, lifestyle adjustments, and medications for those already diagnosed with enuresis. This suggests that individuals in Korea may be more reluctant to visit a doctor for enuresis and prefer to seek anonymous help online.
Psychological factors triggered by social events or cultural differences: The prominence of the relevance to crimes or traumatic incidence topic on Reddit highlights that enuresis may not solely be a result of physical or developmental issues but could also be influenced by psychological factors triggered by social events or cultural differences.
Analysis of the Community Names in Online Platforms
Results of categorization and frequency visualization for community names
In both platforms, users can participate in and upload their posts in the Cafe or Subreddit that are organized around specific interests. This paper aimed to indirectly assess which topic categories users tend to upload enuresis-related posts to, as well as the type of community they expect to receive or provide enuresis-related information. To achieve this, this study collected the keywords used in the names of the Cafes and Subreddits and categorized them based on their thematic nature. The frequency of these categories was then visualized through bar graphs as shown in Figs. 5 and 6.
In Naver Cafe, enuresis-related posts were most frequently found in region-based communities including those centered around locations such as ‘Incheon,’ ‘Paju,’ and ‘Busan.’ The second most frequent posts appeared in general social communities. Notably, there was significant activity in a community specifically for mothers known as MOM Cafe with keywords such as ‘mom,’ ‘mothers,’ ‘moms-holic,’ etc. Additionally, enuresis-related posts were common in communities focused on health and illness concerns including ‘hospital,’ ‘vasectomy,’ ‘breast cancer,’ ‘menstrual cramps,’ ‘vaginitis,’ ‘cystitis,’ and ‘erectile dysfunction.’
In contrast, on Reddit in the US, enuresis-related posts were predominantly found in disease-related communities with keywords such as ‘Adult Bedwetting,’ ‘Prostatitis,’ ‘Interstitial Cystitis,’ ‘Sleep Apnea,’ ‘Pelvic Floor,’ and ‘Insomnia.’ These was followed by hobby-related communities centered around topics like ‘AI Art,’ ‘Netflix,’ ‘Digital Game,’ ‘Stephen King’ as well as academic communities focused on areas such as College Homework, Grammar, Medical School, Scholar, and Scientific Nutrition. In addition, posts also appeared in social communities, particularly those focused on advice, with the keywords including Advice, Off My Chest, Ask Reddit, Relationship Advice.
Analyzing the online spaces where enuresis-related discussions take place reveals that both in South Korea and the US, hobbies and social interactions including counseling are common contexts for discussing the discomfort associated with enuresis. This underscores the fact that enuresis is a condition people are hesitant to discuss openly, especially with strangers. A notable difference is that while US users often seek out disease and medical communities for expert knowledge and medical information, South Korean users tend to rely on local groups or mother-focused communities for information about enuresis. This indicates a greater reliance on word-of-mouth and community advice in Korea, rather than official medical information from urology experts.
DISCUSSION
Overall, online platforms such as Naver Cafe in South Korea and Reddit in the US serve as open and active spaces for discussions about enuresis. Users on both platforms are eager to document and share their experiences with enuresis symptoms, in particular as adults, and are most likely to post in communities focused on disease-related, pediatric caregiving, and womenrelated topics.
In both countries, there is a tendency to discuss symptoms in children online to seek advice, with South Koreans being more inclined to talk about pediatric enuresis. On Reddit, community titles that can infer gender are more likely to include words like ‘mom’ and ‘woman’ rather than men. This suggests that women are more likely to be the authors of these posts.
Key differences between the 2 platforms include the followings. In South Korea, there is a stronger tendency to seek information and share experiences before visiting a doctor and receiving an accurate diagnosis. In 4 out of 5 topics including topics 1, 3, 4, and 5, keywords extracted from posts frequently focused on describing symptoms and seeking advice. Many of these posts were found in the communities related to local areas, parenting, women, and family, where users sought practical advice on diagnosis, symptoms, and hospital recommendations. There was also a noticeable preference for traditional media treatments. This suggests not only a stronger negative perception of enuresis but also a tendency to rely on alternative treatments to address fears of visiting a urologist or to obtain information from laypeople rather than medical professionals.
In contrast, online communities in the US tend to focus on sharing personal experiences and emotional states during treatment and management after diagnosis, particularly in diseaserelated spaces. Users are more likely to discuss their own symptoms rather than express concern for family members. However, the prevalence of enuresis-related posts in communities focused on counseling indicates that enuresis remains a difficult and stigmatized condition to disclose.
A significant cultural difference observed in the Reddit compared to the South Korean data is the presence of topics related to crime and lifestyle modification. Additionally, all of the remaining topics except for topic 1 frequently featured specific disease and medication. This suggests that users on Raddit, who have already been diagnosed or are undergoing treatment, often discuss and share detailed information about the condition, more so than users in South Korea. As a result, there was a stronger tendency to manage enuresis in daily life including lifestyle and dietary adjustments.
In conclusion, in this study, data from online community posts in South Korea and the US were collected and analyzed using text-mining techniques to better understand public perceptions of enuresis, a condition that is often difficult for patients to discuss openly. The results revealed notable differences in how Korean and American community users perceive enuresis symptoms and treatment.
In Korea, people tend to be very cautious about choosing and visiting hospitals, and often to prepare for the visit with the help of others. This suggests a need for medical institutions to lower the barriers to access and foster closer, more approachable relationships with patients to prevent adverse effects of relying on nonprofessional opinions. In addition, since female family members appear to be overwhelmingly involved in pediatric visits, healthcare organizations could benefit from targeting married women in their marketing strategies. Compared to the US patients, South Koreans are significantly less likely to discuss their own illnesses or symptoms online, so there is a need to raise awareness to encourage adults to recognize and address their symptoms in a timely manner. Given the tendency for individuals to seek advice and support from others with similar experiences rather than immediately consulting a specialist, it is crucial to ensure that accurate information and effective coping strategies are widely disseminated.
This study is meaningful in that it diagnoses public coping methods and perceptions of enuresis—an often-taboo subject in offline discussions—using big data and text-mining techniques. The research also highlights the ability to identify differences in responses to enuresis across countries without being constrained by physical distance and provides valuable insights into cultural attitudes toward this condition. Furthermore, this methodology offers a foundation for examining perceptions and attitudes toward various stigmatized diseases and for developing guidelines for appropriate initial responses.
The study has limitations inherent in the big data analysis approach. While big data analysis can identify correlations between variables, it does not necessarily establish a direct causal relationship between them [23]. This can lead to the potential pitfall of conflating correlation with causation when interpreting the findings. In addition, the data was collected from only 2 online communities with distinct characteristics, and the analysis was also based on a dataset obtained through the official APIs provided by these platforms. In this regard, further studies can be done with a larger quantitative sample and more diverse communities. Also, it can be applied to other countries in order to identify cultural differences in how difficult-to-discuss disease are treated. Such an approach could significantly contribute to improving awareness and disseminating accurate information about these stigmatized conditions within diverse cultural contexts.
PDF Links
PubReader
ePub Link
XML Download
Full text via DOI
Download Citation
Print
