Download PDF
Case Studies > Health-Links Success Story: Arabic Sentiment Analysis Solution

Health-Links Success Story: Arabic Sentiment Analysis Solution

Technology Category
  • Analytics & Modeling - Machine Learning
  • Analytics & Modeling - Predictive Analytics
  • Application Infrastructure & Middleware - Data Exchange & Integration
Applicable Industries
  • Healthcare & Hospitals
Applicable Functions
  • Business Operation
  • Quality Assurance
Services
  • Software Design & Engineering Services
  • System Integration
The Challenge
Health-Links faced several challenges in analyzing patient and employee feedback. The primary issue was the need for a robust, cloud-based sentiment analysis solution that could natively understand Arabic and handle mixed Arabic-English responses. The client needed to analyze a massive amount of unstructured text from over 12 million surveys annually. Manual interpretation was prone to human bias and errors, leading to inaccurate insights. The client required a solution that could provide sentiment scores for themes specified in the Saudi Complaints Taxonomy, enabling healthcare organizations to understand the pros and cons of their operations.
About The Customer
Health-Links is a specialized healthcare consultancy based in Jeddah, Saudi Arabia. The company partners with the Ministry of Health, KSA, and other healthcare leaders to improve the quality of healthcare in the Gulf region. Health-Links focuses on identifying gaps in care services and mapping the patient journey within hospitals using data-backed insights. The company collaborates with Press Ganey Associates for inpatient experience measurement, performance analytics, and strategic advisory solutions. Operating throughout the Middle East, Health-Links leverages Press Ganey’s solutions under a localized model to enhance healthcare delivery.
The Solution
Repustate provided Health-Links with a customized sentiment analysis solution capable of handling hundreds of API calls in seconds. The cloud-based solution offered aspect-based sentiment scores for each topic specified by the KSA Ministry of Health. The model was designed to natively analyze Arabic text without relying on English translations, classifying comments into Positive, Negative, Neutral, and Mixed categories. Repustate trained the model according to the Complaints Taxonomy and discovered new relevant topics during development. The model was reviewed and re-trained based on feedback from the Ministry, achieving an accuracy rate of 81% within a month and three iterations.
Operational Impact
  • Health-Links can now measure performance and drive improvements for safe, high-quality, patient-centered care.
  • The solution helps Health-Links manage continuously growing data and discover trends from historical data.
  • Granular aspect-themed sentiment scoring aids the Ministry of Health and healthcare organizations in prioritizing policy decisions.
Quantitative Benefit
  • The sentiment analysis model achieved an accuracy rate of 81% within a month and three iterations.
  • Health-Links conducts more than 12 million surveys annually, all of which are now analyzed using the automated solution.

Related Case Studies.

Contact us

Let's talk!

* Required
* Required
* Required
* Invalid email address
By submitting this form, you agree that IoT ONE may contact you with insights and marketing messaging.
No thanks, I don't want to receive any marketing emails from IoT ONE.
Submit

Thank you for your message!
We will contact you soon.