This paper focuses on the evaluation and recommendation of healthcare applications in the mHealth field. The increase in the use of health applications, supported by an expanding mHealth market, highlights the importance of this research. In this study, a data set including app descriptions, ratings, reviews, and other relevant attributes from various health app platforms was selected. The main goal was to design a recommendation system that leverages app attributes, especially descriptions, to provide users with relevant contextual suggestions. A comprehensive pre-processing regime was carried out, including one-hot encoding, standardisation, and feature engineering. The feature, “Rating_Reviews”, was introduced to capture the cumulative influence of ratings and reviews. The variable ‘Category’ was chosen as a target to discern different health contexts such as ‘Weight loss’ and ‘Medical’. Various machine learning and deep learning models were evaluated, from the baseline Random Forest Classifier to the sophisticated BERT model. The results highlighted the efficiency of transfer learning, especially BERT, which achieved an accuracy of approximately 90% after hyperparameter tuning. A final recommendation system was designed, which uses cosine similarity to rank apps based on their relevance to user queries.

Here are a few use cases for this project:

1. Smart Vending Machines: The model can be used in smart vending machines to accurately identify and provide inventory information about available snacks, helping with automatic restocking alerts.

2. Retail Store Automation: The model can be used for checkout-free shopping experiences similar to the Amazon Go. Shoppers can simply take the snacks they want, and the computer vision would identify it and charge them automatically.

3. Dietary Apps: The model can be integrated in nutrition and diet apps to help users identify their snack's nutrition information just by taking a picture, making it easier to maintain a healthy diet.

4. Waste Sorting: The model can help differentiate packaging materials used for different snacks, aiding in waste sorting and recycling efforts at an industrial scale.

5. Snack Manufacturing Quality Control: The model can be used on the production line to identify misprints or mispackaging issues ensuring that every package is correctly labeled for sale.

Here are a few use cases for this project:

1. Bakery Inventory Management: Sdoba can be used by bakery owners to track the inventory levels of various pastry items in real-time by identifying and counting them in images taken within the store or production area. This helps to manage stock effectively and plan for replenishments.

2. Automatic Billing System in Cafes/Canteens: The computer vision model can be integrated into a Point of Sale (POS) system to automatically recognize and bill customers for their selected pastries. This helps to speed up the checkout process and reduce manual errors during billing.

3. Quality Control in Production: Sdoba can be leveraged by pastry manufacturers to identify and inspect the quality of their products, including shape, uniformity, and appearance. Images taken during the production process can be analyzed to ensure the products meet the desired standards and immediately flag any inconsistencies.

4. Recipe App Integration: By using Sdoba, recipe apps can recognize and suggest corresponding recipes for users based on the images of the pastries they upload. This makes it easier for users to find and recreate their favorite pastry dishes at home.

5. Nutritional Analysis Assistance: The model can be further developed to estimate the nutritional content (calories, fat, etc.) of each pastry class recognized in the images, allowing diet tracking apps or smart restaurant menus to provide more accurate nutritional information to users based on their selected pastries.