Sentiment analysis on autism spectrum disorder using twitter data

Abstract

Autism is the behavioral disorder; it leads to developmental disorder and repetitive disorder. Autism Spectrum Disorder (ASD) is one of the disorders of autism. ASD diagnosed based on assessing social behavior and clinical tests of autistic children. The assessment process is conducted by multi disciplinary team of doctors based on the universal standard questionnaire. The objective of this paper is to predict analytical values from semi structured data posted in twitter by individuals, caregivers of autistic children. The different Natural Language Processing and Topic Modeling algorithms are applied to analyze autism based on tweets collected. Approximately 10k tweets dataset is used for this analysis. NLP and topic modeling are reliable and efficient methods to perform text analysis with 50% less time and the results are 90% accurate compared to regular text processing methods. The analysis performed for genetic analysis, effect of vaccination analysis and behavior analysis. The analytical results are used to learn the genetic impact on ASD, vaccination effect on ASD. And also used learn the behavior changes and population of autistic children. Results are useful for the parents, caregivers, individuals and other researchers to learn about ASD.