Usability Evaluation for Mobile-Based Nutritional Food Recommender System

. COVID-19 was designated as a global pandemic by WHO and Government Regulation of the Republic of Indonesia Number 21 of 2020 concerning Large-Scale Social Restrictions for the handling of Corona Virus Disease 2019 (COVID-19). Eating healthy foods with balanced nutrition can increase the body's immunity during a COVID-19 pandemic. Since many foods sold freely are not guaranteed in nutritional ingredients, cooking becomes a better alternative . Previous research has implemented the Simple Additive Weighting (SAW) as a recommendation algorithm for recipes recommendation and gives significant result. It has not tested the usability of the system when it is used by end users. This research improved data and variable from previous research so that they are suitable to be applied to support the nutritional adequacy during COVID-19 pandemic. The result of usability testing shows that there is a significant increase in SUS score from 63.5 to 87.


Introduction
Corona Virus Disease 2019 (COVID-19) has become a global pandemic since the end of 2019, therefore, to solve this problem, the government of Indonesia made Government Regulation of the Republic of Indonesia Number 21 of 2020.The policy makes people do physical distancing and work from home to help suppress the spread of COVID-19.However, by implementing this policy, the community's lifestyle will potentially become a sedentary lifestyle which will affect the decrease in physical activity because it decreases the exercise habit [1,2].
The decrease in exercise habits will cause the decrease of immunity of the body.Therefore, in addition to do physically distancing to limit the spread of COVID-19, protection from inside, which is increasing the body's immunity is needed.
Healthy food is all food containing balanced nutrition.Balanced nutrition is food that has three important elements in the human body, namely energy substances, building blocks, and regulatory substances, as well as maintaining food portions adjusted to the body's ability to accommodate food [3,4].For making healthy food, recipes will be needed as a reference in processing the food, which will facilitate the cooking process [5,6].Based on the questionnaire with 10 respondents that active in searching food recipes, 8 of them stated that in searching for recipes they often faced many recipe choices, in this case, it will cause a longer duration to start the cooking process because they will compare one recipe with another recipe.The 10 respondents in background of studies are different with the 5 respondents in the usability testing.
To overcome these problems, the authors offer a solution in the form of a system that can support decisions from the problem of finding healthy recipes using the Simple Additive Weighting method or can be shortened as SAW.Recipe recommendation system has been developed previously [7,8,9,10,11].SAW assigns a value to each criterion which is then summed and the final decision results is the recommendation [13,14,15,16,17].The selection of the SAW method was based on previous research [18].The rank consistency testing in SAW-based recipe recommendation system is 100% consistent, it means there is no rank reversal in [18].
Previous study [18] has not tested the usability of the system when it is used by end users.The data and variables in [18] are still general, so this research improved data and variable so that they are suitable to be applied to support the nutritional adequacy during COVID-19 pandemic as can be seen in Table 3.We conducted a user study to measure the usability of the proposed system.The usability testing will use the SUS (System Usability Scale) questionnaire by comparing the results of the SUS scores from previous research and current research.

Research Method
Research methods of this study is described in figure 1. Literature study is a reference search that contains an explanation of the theoretical basis and several previous research that can support this research.The final result of SAW is recommendations for healthy recipe from the first to the last ranking.
We conducted a user study to measure the usability of the proposed system with the SUS (System Usability Scale) questionnaire.Usability test is conducted to understand user's perception about the application.The SUS test is a questionnaire to evaluate usability of a system.It was discovered by John Brooke [19] and it is cheaper and quicker than any usability test.The tests in this research have conducted by comparing the results of the SUS scores from previous research and this research [20].
SUS questionnaire is given to 5 respondents.Testing with 5 people let us find almost as many usability problems as you'd find using many more test participants [21].There are 10 questions to be answered by 5 end users of the application.Each question has a scale of 1 to 5, which means 1 for strongly disagree and 5 very agree The SUS question is [19]: 1.I think that I would like to use this system frequently.2. I found the system unnecessarily complex.3. I thought the system was easy to use. 4. I think that I would need the support of a technical person to be able to use this system.5.I found the various functions in this system were well integrated.6.I thought there was too much inconsistency in this system.7. I would imagine that most people would learn to use this system very quickly.8.I found the system very cumbersome to use. 9.I felt very confident using the system.10.I needed to learn a lot of things before I could get going with this system.For each question with an odd number, the existing score is subtracted by number 1.For each question with an even number, the number 5 will be subtracted from the existing score.Then the sum of results of points 2 and 3 are multiplied by number 2.5 as can be seen in Equation 1.
After getting the final score, the next step is to determine the grade of the assessment results based on acceptability, grade scale, and adjective rating [22] as shown in Figure 2. Fig. 2. Analysis of SUS score [22] 3 Result and Analysis Data acquisition is conducted by taking data from books, the internet, and calorie calculation sites.This study improves data from previous research [18] as can be shown in Table 1.Variable of the previous research are time, calorie and the number of steps.The data from the recipes that used in this research are in Table 2.This study removes time as criteria that used in previous study and replace it with the quantity of ingredients because this variable is more measurable and the cooking time for each person can't be the same.This study also added nutritional information on each recipe as a new feature to improve user experience.The difference between previous and current research can be seen in Table 3.The user interface for the recipe recommendation system will be 4 pages consisting of a criteria input page, a recipe recommendation page, a recipe detail page, and the page of food nutrition details can be seen in figure 2.
On the first page, the user enters the value of the time variable and the level of complexity, then press the continue button, so a recommendation will appear.
After the application is developed, the usability testing process is carried out using the SUS (System Usability Scale) questionnaire with 5 respondents who are the end users of the application and they are between 20-40 years old.
After getting data from 5 respondents, the data will be processed using SUS calculation testing.The results of the SUS questionnaire in previous research [18] is in Table 4 and 5.   Results of the SUS testing with 5 respondents in previous research, the score was 63.5.So, we continue SUS testing for this research.The responses from respondents in this current study with SUS testing questions can be seen in Table 6.

Conclusion
We conducted a user study to measure the usability of the proposed system.This research improved previous research in usability aspect from end users.The result of SUS score from 63.5 to 87.The previous research has Grade D on the Grade Scale; High Marginal Ranges on the Acceptance ranges; and OK on the Adjective Rating.This research has 87 in SUS score so it means that it is in Grade B on the Grade Scale; Acceptable on the Acceptability Ranges; and Excellent on the Adjective Rating.For future works, it is suggested to add information of anti-nutrients in every recipe.

Fig. 1 .
Fig. 1.The research method diagram used to describe the entire research flow

Fig. 3 .
Fig. 3.The Interface of implementation with upgrading application.

Table 1
Recipe Data From Previous Research

Table 3
Differences between previous research and current research

Table 4
Previous research respondents response

Table 5
The results of the calculation SUS testing of the previous research p-ISSN: 2540-9433; e-ISSN: 2540-9824

Table 7
The results of the calculation SUS testing of the current research This research SUS score is 87.It means that it can be categorized in Grade B on the Grade Scale; Acceptable on the Acceptability Ranges; and Excellent on the Adjective Rating