User:Neema5150

Fitness and Mental Health on Social Media
Twitter has been widely used in the data science field due to its massive, publicly accessible data source. It has a broad range of applications from economics to health. Twitter can be used to predict political outcomes, such as the elections, and try to explain those outcomes. It can also be used to measure the effectiveness of advertising to study economic behavior. Identification of emergency events has also been successful through Twitter, in addition to relaying this information to affected communities and authoritative agencies. It has also been used in the health field to track influenza by identifying flu related tweets. To study the relationship between language use and psychology, Twitter has been used to develop classifiers of human affective states in social media. Additionally, Twitter can serve as a platform to discuss socially stigmatic issues, such as mental health issues, which can be used to study emotional health. Aside from Twitter, other social media streams, such as Reddit, can act as a sharing community, allowing people to exchange information regarding health challenges. Twitter and Reddit have been used in studies to understand motivation behind fitness, how it provides social support for mental illnesses, and to model mood public mood.


 * 1) Twitter and Fitness
 * 2) Mental Health on Reddit
 * 3) Mood and Emotions on Twitter

Fitness through Twitter
Maintenance of healthy behavior can be challenging compared to initiating a healthy behavior (e.g. workout beginners). Although social support through social media, such as Twitter, has shown weight loss the content and motivation behind messages on social media that promote ongoing health behavior is unclear. This is important as many people tend to relapse back to their unhealthy lifestyle shortly after starting a healthy routine. Qualitative analyses of health/fitness related tweets and semi-structured interviews of twitter users can help understand the messages that motivate health maintenance. Tweets are considered to be unstructured data, so in order to analyze them they need to be converted into structured data. This can be done through an affinity analysis that can gather common themes and group the tweets. Same goes for interviews where they are analyzed through constant comparison method, axial coding and selective coding.

1,000 random Twitter posts from the Twitter search API that included posts with the words gym, workout, calories, diet, weight, and/or health showed that health/fitness related tweets contain actualization and sentiment. Actualization is when people write about their plans and goals, post about the achievements, and actions they avoided. The sentiment aspect tends to capture the emotions expressed in the health-related posts. These could either be positive or detrimental sentiments towards the lifestyle choices. Also, 12 daily health bloggers on Twitter indicated that lurking (learning by observation), the need for a lifestyle compatibility, the feeling of accountability to an audience, and navigating through feedback is what enabled motivation and healthy practice. Accountability to an audience (followers) encourages maintenance behaviors as participants feel a sense of responsibility to share any sort of health progress. This encourages users to follow through with their workout. Overall though, Twitter allows users to engage in weight maintenance activities feasibly and in a natural way.

Mental Health on Reddit
Mental illness is surrounded by stigma making it hard for people who suffer to cope with the stress, pain, and the condition. Certain social media streams, such as reddit, have become a sharing community allowing people to exchange information regarding health challenges. Seeking and sharing information online has shown to be effective in helping people cope with their problems, therefore it is important to understand self-disclosure in mental illness communities and examine factors that contribute to social support online.

20,411 reddit posts and 97,661 comments, collected from reddit's official API, showed that anonymity allowed greater self-disclosure around stigmatic topics and allowed prudent, open communication. Reddit posts and comments are unstructured data. The linguistic attributes manifested in the posts and comments can be examined using psycholinguistic lexicon LIWC, and comments are best analyzed when clustered into different topics using LDA. Social support on reddit is driven by lowered inhibition postings, which are more self-attention focused and discuss relationship and health issues. This seemed to gather greater community support through comments. Social support can come in many different forms though: emotional, instrumental, information, and prescriptive advice.

Mood and Emotion on Twitter
Microblogging, or tweeting, has become increasingly popular where people broadcast their daily activities, conversations, and share information on new and current affairs. Regardless of the message content, they all convey some mood of state of the author that is either explicit or reflective from the message. Being able to accurately asses people's mood on Twitter due to certain sociocultural events is valuable and important, because it shows how certain phenomena effect people emotionally. Such a task also requires an instrument that can accurately model public emotions (e.g. a psychometric instrument, the profile of mood states (POMS)).

The POMS assigns a particular number to a mood (i.e. mood value), and essentially converts unstructured data, like tweets, into structured data by looking for certain adjectives within a tweet. 9,664,952 tweets, collected during a timeline of major political, cultural, social, economic, and natural events between August 1 and December 20, 2008, that were analyzed by POMS showed the validity of this model in detecting public sentiment and associating fluctuations with socioeconomic events. The POMS model measured six dimensions of mood: Tension, Depression, Anger, Vigor, Fatigue, and Confusion. After converting the mood values into a z-score it was found that events in social, political, cultural, and economic sphere do have a significant effect on dimensions of public mood. Economic events that correspond to rapid changes seem to have a significant effect in public mood.