Toward Automating HIV Identification: Machine Learning for Rapid Identification of HIV-Related Social Media Data

Abstract

Introduction “Social big data” from technologies like social media, wearable devices, and online searches continue to grow and can be used as tools for HIV research. Although researchers can uncover patterns and insights associated with HIV trends and transmission, the review process is time-consuming and resource intensive. Machine learning methods derived from computer science might be used to assist HIV domain experts by learning how to rapidly and accurately identify patterns associated with HIV from a large set of social data. Methods Using an existing social media dataset that was associated with HIV and coded by an HIV domain expert, we tested whether four commonly used machine learning methods could learn the patterns associated with HIV risk behavior. We used the 10-fold cross validation method to examine the speed and accuracy of these models in applying that knowledge to detect HIV content in social media data. Results and Discussion Logistic regression and random forest resulted in the highest accuracy in detecting HIV-related social data (85.3%) while the Ridge Regression Classifier resulted in the lowest accuracy. Logistic regression yielded the fastest processing time (16.98 seconds). Conclusion Machine learning can enable social big data to become a new and important tool in HIV research, helping to create a new field of “digital HIV epidemiology.” If a domain expert can identify patterns in social data associated with HIV risk or HIV transmission, machine learning models could quickly and accurately learn those associations and identify potential HIV patterns in large social datasets.

<p>On the Role of Artificial Intelligence in Genomics to Enhance Precision Medicine</p>

“…where several studies such as the one carried out by Li et al assessed the risk of heart attack utilizing an artificial neuron network. 22 Other studies employed machine learning in order to identify the risk of developing other diseases, such as colorectal cancer, 23 respiratory virus affinity, 24 melanoma, 25 mortality in smokers, 26 depression 1 or HIV transmission, 27 among others.…”

Feature Selection Methods for Mining Social Media

“…This work supported the utilization of social media data for improved messaging in campaigns. Sean D. Young et.al., [20] in their research work mined Twitter data to identify and track risk behaviours of HIV. They tracked the location in which the HIV risk behaviours tweets occurred.…”

“…The SVM classification can be applied to both binary and multiple classification [20]. For multi-classification two SVM methods can be used, OVA: One versus All method is used to fit the k class as, , = 1, … .…”

Monitoring HIV Testing and Pre-Exposure Prophylaxis Information Seeking By Combining Digital and Traditional Data

“…For instance, the most recent data for HIV testing on AIDSvu.org is from 2016 and the most recent AIDSvu.org data on Pre-Exposure Prophylaxis (PrEP) usage is from 2018 [2]. These limitations have driven public health to increasingly turn to digital data, such as news, social media, and internet searches, to learn how people seek HIV information [3][4][5][6]. For example, internet search trends can be used to investigate public interests as evident by actor Charlie Sheen's HIV positive disclosure concurring with record levels of Google searches for HIV awareness, HIV testing, and condoms [7].…”