使用社交媒体的健康促进活动:推特主题标签的关联规则挖掘与共现网络分析

Health promotion campaigns using social media: association rules mining and co-occurrence network analysis of Twitter hashtags.

作者信息

Ghahramani Atousa, Prokofieva Maria, de Courten Maximilian Pangratius

机构信息

Institute for Sustainable Industries & Liveable Cities, Business School, Victoria University, 370 Little Lonsdale St, Melbourne VIC 3000, Australia.

Victoria University, Institute for Health and Sport and Australian Health Policy Collaboration, Melbourne, Australia.

出版信息

BMC Public Health. 2025 Jan 7;25(1):67. doi: 10.1186/s12889-024-21255-5.

Abstract

BACKGROUND

Social media hashtags play a significant role in increasing the visibility of health information by making it easier for people to explore health-related content. Health promotion campaigns use campaign-specific hashtags to disseminate health-related messages, enabling individuals to access accurate and timely resources and updates. The study aims to discover patterns of connection between hashtags and identify the most influential hashtags used on Twitter in the American Heart Month campaigns.

METHOD

We collected a total of 73,288 tweets containing #AmericanHeartMonth between January 2019 and March 2023 and retrieved 18,143 original tweets, 42,930 retweets, 2,519 quotes, and 20,846 likes related to the past five campaigns. We adapted co-occurrence network analysis to explore the patterns of relationships between hashtags and association rules mining to assess the quality and strength of association between the co-occurred hashtags.

RESULT

While #AmericanHeartMonth, #OurHearts, and #HeartMonth play central roles in all hashtag co-occurrence networks, the results of association rules mining indicate a significant association of #OurHearts within the networks. The highest density of hashtags has been observed in the quoted tweets, introducing a new range of hashtags such as #GoRedForWomen, #WearRedDay, #HeartDisease, and #HeartHealth by Twitter users, indicating the positive correlation between co-occurring hashtags and users' engagement. The results of quality measurements of association rules (Lift > 1) indicate positive relationships between the co-occurred hashtags in the top 5 rules in all data subsets.

CONCLUSION

We employed co-occurrence network analysis and association rules mining as powerful techniques to identify influential hashtags that may have a central role in health-related discussions and drive engagement within the co-occurrence hashtag network. In conclusion, we recommend additional hashtag structures in conjunction with heart health-related topics to improve community building and the effectiveness of disseminating messages in future heart health promotion campaigns. The study contributes to knowledge and practice by offering a structured and data-driven approach and providing practical guidance for public health practitioners, professionals, and organisations to optimise content, targeting, and messaging to reach and engage a broader audience with health-related information.

摘要

背景

社交媒体标签通过使人们更容易浏览与健康相关的内容,在提高健康信息的可见性方面发挥着重要作用。健康促进活动使用特定活动的标签来传播与健康相关的信息,使个人能够获取准确及时的资源和更新。本研究旨在发现标签之间的关联模式,并识别美国心脏月活动中在推特上使用的最具影响力的标签。

方法

我们收集了2019年1月至2023年3月期间共73288条包含#美国心脏月的推文,并检索到18143条原创推文、42930条转发推文、2519条引用推文以及与过去五次活动相关的20846条点赞推文。我们采用共现网络分析来探索标签之间的关系模式,并运用关联规则挖掘来评估共现标签之间关联的质量和强度。

结果

虽然#美国心脏月、#我们的心脏和#心脏月在所有标签共现网络中发挥核心作用,但关联规则挖掘结果表明#我们的心脏在网络中存在显著关联。在引用推文中观察到最高的标签密度,推特用户引入了一系列新的标签,如#为女性穿红衣、#穿红衣日、#心脏病和#心脏健康,表明共现标签与用户参与度之间存在正相关。关联规则质量测量结果(提升度>1)表明,在所有数据子集中前5条规则中共现标签之间存在正相关关系。

结论

我们采用共现网络分析和关联规则挖掘作为强大的技术来识别有影响力的标签,这些标签可能在与健康相关的讨论中发挥核心作用,并推动共现标签网络中的参与度。总之,我们建议结合与心脏健康相关的主题采用额外的标签结构,以改善社区建设,并在未来的心脏健康促进活动中提高信息传播的有效性。该研究通过提供一种结构化的数据驱动方法,并为公共卫生从业者、专业人员和组织提供实用指导,以优化内容、目标定位和信息传递,从而接触更广泛的受众并使其参与与健康相关的信息,为知识和实践做出了贡献。

https://cdn.ncbi.nlm.nih.gov/pmc/blobs/4000/11706027/9b4960915667/12889_2024_21255_Fig1_HTML.jpg

文献AI研究员

20分钟写一篇综述,助力文献阅读效率提升50倍

立即体验

用中文搜PubMed

大模型驱动的PubMed中文搜索引擎

马上搜索