*2.9. IoT Enabled System Architecture*

This system is regarded as a multiple-user access system, allowing numerous users to connect to the cloud simultaneously, as shown in Figure 3. There is only a single universal receiver shared by all users. An IoT system with cloud administration was created to classify obesity. Because it is a distributed system, the cloud is the best solution for a healthcare system that enables doctors to obtain data more easily. Our suggested IoT system comprises four key phases: (1) data collecting, (2) textual data classification, (3) diagnosis, and (4) user interface. Its goal is to lower disease rates through early detection of obesity.

**Figure 3.** Proposed IoT system architecture.

This figure demonstrates that the user is the origin of the entire process. With the web application interface, users engaging with the server and application interfaces are directly coupled. Therefore, when a user interacts with the web interface, a specific request is sent to the server. Upon receiving a request, the server examines it to determine what is the need of the user (obesity prediction, check his/her history, download report, or doctor's advice). Then the server will decide where to transmit the user's request after considering the needs of the user. Therefore, the server looks for an expert system that can handle the user's request and deliver the results. The server assigns the user's duties after identifying the optimal expert system. The user's task is inputted into the expert system as a string because the entire model is reliant on textual information, which is utilized to identify obesity in its early stages. After receiving a string input, the algorithm eliminates any extraneous words that are found during the prediction stage. After eliminating superfluous words, the user-provided data are used by the prediction engine to make predictions. Following the calculation of the outcome, the results are sent to the expert system. The server receives the results that the expert system collects. After obtaining the expert system's results, the server sends it to the web interface, where the user can access his/her results and move forward in light of the report.
