Senior Design Projects
ECS193 A/B Winter & Spring 2022
Leveraging Social Media Data to Inform Family Caregiving Research
| ********** | |
| Yong K. Choi | |
| UC Davis Public Health Sciences |
Project's details
| Leveraging Social Media Data to Inform Family Caregiving Research | |
| The exponential growth of social media has led to widespread of social media use in healthcare. The social media platforms encourage family caregivers to be a part of an online community and share their personal stories with others to receive support, share tips and advice and feel less alone. The anonymous nature of the platform could allow the users to freely share their thoughts and needs that they might not feel confident or comfortable expressing in real-life settings. Previous research show that social media platforms are now more commonly incorporated in the decision-making process to aid caregivers with making informed decisions regarding their loved one’s care. Therefore, social media data can provide valuable insight into the caregiver’s concerns and priorities without the financial, temporal, and spatial encumbrances of more traditional community-engagement methods. While these newer methods cannot replace the more traditional ones, social media data analysis of caregiver posts may prove a useful starting point for developing relevant and effective family-centered interventions. | |
| Aim 1: Create a data repository of family caregiver online discussion forums The first aim is to create a data repository of family caregiver online discussion forums to enhance investigator capacity to conduct family caregiving research. We recognize the importance of the ‘open science’ agenda to make research data to be accessed, used, and shared by other investigators for secondary analysis. We would like to make this data available to the public to create opportunity for collaboration and innovation while increasing scientific and reproducibility in family caregiving research across various disciplines. Aim 2: Conduct exploratory ML/NLP analysis The second aim is to conduct exploratory ML/NLP analysis to do topic modeling and sentiment analysis to analyze the data gathered in Aim 1. The research team has created a potential list of research questions we would like to explore with the data. | |
| 1) A Family Caregiver Social Media Corpus: A student will work with the client to scrape the data from publicly available caregiver forums listed above. We would like to start with the following two data sources listed below. 1. Alzheimer's disease and related dementias (AD/ADRD) caregiver forum a. Link: https://www.alzconnected.org/discussion.aspx?g=topics&f=151 A free online forum run by the Alzheimer’s Association in the United States, offers a range of forums for people with ADRD and their caregivers or friends and family members. 2. Amyotrophic Lateral Sclerosis (ALS) caregiver forum a. link: https://www.alsforums.com/community/forums A free online community for persons affected by amyotrophic lateral sclerosis and motor neuron disease. 2) Exploratory ML/NLP analysis: A student group will work with the client and UC Davis DataLab to perform exploratory topic modeling and sentiment analysis with the data corpus collected in Aim 1. The example exploratory research questions are listed below: ● What are the unmet needs of family caregivers caring for individuals with Alzheimer’s and related dementia? . . . caring for individuals with ALS? ● Do these needs differ for those engaged in complex care including performing medical/nursing tasks in the home? ● How do family caregivers describe their own health status? ● . . . physical health status including visits to the ED, hospital admissions ● . . . mental health status including depression, anxiety and loneliness? ● . . . social health status including relationships with the care recipient, friendships and social isolation? ● . . . financial hardship? ● . . . spiritual health? | |
| 1) Webscraping with Python: https://realpython.com/beautiful-soup-web-scraper-python/ 2) Topic modeling with Python: https://towardsdatascience.com/end-to-end-topic-modeling-in-python-latent-dirichlet-allocation-lda-35ce4ed6b3e0 | |
| ********** | |
| 30-60 min weekly or more | |
| Open source project | |
| Attachment | N/A |
| Yes | |
| Team members | N/A |
| Tyler Mayxonesing | |
| N/A |