Data science , through its machine learning algorithms , also provides valuable insights and accurate predictive models and assists companies in making decisions .

To fulfil halal certification requirements , data science evaluates the gap between the status quo of the companies against the halal requirements , improves how the supply chain works , gives customers a better experience and ensures the halal assurance system is in place . Allah said in the Quran :

“ And give full measure when you measure , and weigh with an even balance . That is the best [ way ] and best in result ” [ Q . Al-Isra ’, 17 : 35 ]

Allah has instructed us that we carry out activities with precision and accuracy . More so , practising precision and accuracy in matters involving halal products and services is paramount . This article examines four ways in which data science is used in the halal industry in the following areas :

• halal certification ,

• consumer opinions and insights ,

• supply chain optimisation , and

• food fraud

HALAL CERTIFICATION

In halal certification , companies must show that their ingredients and manufacturing processes comply with Islamic law . Five halal supply chain components in halal-certified products are routinely checked : product ingredients , production , packaging , storage , and transportation .

This certification process can be lengthy and costly since it requires physical inspections and paperwork .

So , how can data science assist in halal certification ? The application of data science as a verification tool assists in certifying halal products . Based on the chemical composition of ingredients like alcohol , pork , or gelatin , companies could train machine learning algorithms to spot these non-halal ingredients in products .

These algorithms evaluate big data from databases with information about suppliers , production processes , and reports from laboratory testing to ensure the ingredients used meet the halal standards .

Halal Certification Bodies ( HCBs ) can adopt this initiative to collate databases from all companies applying for halal certification and divide the database into successful versus rejected applications . Further investigation of the rejected applications may provide details on the potential halal toyyiban threats from various sources .

This includes incomplete information on the ingredient status , doubtful ingredient and flow of the manufacturing process , fake halal certificate issuance and halal certificate issuance from unrecognised HCB .

Using data science in this sense would prevent the recurrence of non-halal products claimed to be halal . Such information , like the issue of meat cartels , can be added as part of the database to enable data science to identify the possibility of its recurrence .

BY MUHAMAD SHIRWAN ABDULLAH SANI

AND

NOOR FAIZUL HADRY NORDIN

INTERNATIONAL INSTITUTE FOR HALAL RESEARCH AND TRAINING ( INHART ) INTERNATIONAL ISLAMIC UNIVERSITY MALAYSIA

CONSUMER OPINIONS AND INSIGHTS

Companies must understand people ’ s behaviours and preferences to create effective marketing strategies and products that meet customer needs . Data science analyses social media , online reviews , and survey data to understand human behaviour .

Facebook , Twitter , and Instagram are good sources of information because customers can provide feedback and opinions in real-time . Data science could determine halal food and product preferences by analysing hashtags , keywords , and feelings in social media posts .

Reading online reviews can also reveal customer preferences . Data science evaluates reviewers ’ topics , complaints , and ratings to improve products and meet customer needs and demands . Data science analyses survey data like consumers ’ demographics , buying habits , preferences , behaviour and satisfaction to enhance marketing campaigns and customer satisfaction .

However , online reviews collected from shopping applications such as Shoppe and Lazada are preferable to the survey since the former manages information on customer satisfaction after purchase .

The machine learning finds patterns from the reviews and surveys by improving predictive models using regression models , decision trees , and neural networks to predict future consumers purchasing preferences and assists companies ’ decisions and operations .

OPTIMISATION OF THE SUPPLY CHAIN

To ensure on-time , in-tact delivery , companies must prioritise supply chain optimisation . Adopting the application of data science in business operations improves the supply chain by collecting and analysing quality control , logistics and inventory data .

Quality control activity gathers data to improve product quality and reduce contamination in halal-certified products and services . Data science analyses inspections , reports from laboratory testing , and customer

complaints to identify potential halalan toyyiban threats along the supply chain .

This identification involves gathering data during processing , packaging storage and transportation , which covers the traceability requirement outlined in the halal assurance system . The compiled data is saved as a database , and the machine learning algorithm could identify the potential halalan toyyiban threats when unusual information against the database is found ; hence , it suggests the possible root cause .

Companies could proceed with corrective action , verify the root cause of the halalan toyyiban threats and prepare preventive measures to avoid the recurrence . As the database information increases and advanced machine learning algorithms are used , data science may propose corrective and preventive measures for the company ’ s operations .

Data science also assists in optimising transport and delivery times by analysing logistics data , Global Positioning System ( GPS ), traffic conditions , and weather data to find the fastest and cheapest routes . Inventory management data helps maximise stock and minimise waste . Data science also facilitates determining product stock levels by analysing sales , production , and lead times .

Furthermore , data science analyses supply disruptions , natural disasters , and geopolitical risks to help the industry run and minimise losses . This application has been proven to be effective during the Covid-19 pandemic .

Data science supports new technology in the supply chain , like the ‘ digital twins ’ technology , where it provides insights and solutions for different parts of its development and deployment . The digital twin technology replicates a real-world product , system , or process used for simulation , integration , testing , monitoring , and maintenance .

Data science can be used to prepare data from various sources , such as Internet of Things ( IoT ) sensors , historical data , and