We live in a world that contains many inexpensive, artificial and harmful things, and the packaged food industry is no exception. As many humans consume packaged food, the impact this can be significant. To better understand the reality, it is worth watching the above documentary since it is an eye opener and also served as an inspiration to continue our project. It has already impacted many people, but there should be more tools/applications that create similar awareness and that is one motivation for our work.
Summary
You are in a supermarket at cereal aisle looking for healthy breakfast options and you find it difficult to choose one from a wide variety. Finally, you find one that says it contains fruits and it even looks healthy from the colorful fruits' pictures on the box. At this point, you may choose to buy it ignorantly or inspect before buying.
This is exactly where the application based on the Food Provenance ontology can come in handy. Imagine if something can guide you in making wise and healthy grocery shopping decisions. Consider the following two scenarios:
Scenario 1 - Without the application: Since you do not know that the product contains artificial flavors and colors and not the real fruits, you may buy it as it is within your price range and it may appear healthy.
Scenario 2 - With the application: Now that you know the product contains artificial flavors and colors and not the real fruits, and you also know the health hazards associated with these additives, you may choose not buy it unless you want to risk your health. Plus, the application may also show the available healthy substitutions if any.
It may be considered better to buy packaged food free of artificial and harmful ingredients than to risk health hazards and visits to the doctor.
Our aim is to create a Food Provenance ontology that can be used to support questions concerning whether food is safe to consume based on the ingredients, their certifications, and the standards.
Consumers are increasingly concerned about what is in the foods they purchase for consumption. The US diet is largely dependent on processed foods, and the lack of information about these intermediate goods needs to change. Furthermore, concerns over food issues including high amounts of processing, non-labeling of GMOs, additives, and contamination - are high, and answers are not easily found. Consumers may want to be able to easily find this information and use it in planning their meals. In this research, we are interested in evaluating the use of semantic technologies to provide information about the health impacts of ingredients of processed food items found in most supermarkets, so that consumers can make an informed decision about their consumption.
We included existing ontologis such as Bio portal's Food_Ontology and, DBpedia's Food and Ingredient Ontology.
This project is part of the Ontology Engineering 2017 course at RPI. The course web page can be found here.
As part of this course, we developed the Food Provenance Ontology based on the Assignments mentioned in the syllabus. This started with our group having to create a Use Case and Terms List and we refined the scope with the objective of every assignment. The group performance was monitored based on the presentations given in the class and the feedback was used to revise the project further. The exact flow of the ontology can be found in the Conceptual Model and Diagrams. Also, the ontology is leveraged by having SPARQL queries run utilizing the ontology with an objective of having the competency questions answered using the Semantic Technologies. All the work done in this project can be attributed to the complete list of References.
Plus, a poster based on this project won the RPI computer science department Poster Presentation competion held on March 31, 2017.
The work is supervised by Professor Deborah McGuinness and Ms. Elisa Kendall.
To add your suggestions or contributions to the ontology, please see our Getting Involved web page that follows the Maintenance Policies.
Benefit: We encode the semantics of the domain in an ontology, which is written using the World Wide Web Consortium's Recommended ontology language for the web (OWL). The underlying representation language is based on the Description Logics. By doing so, it is easier to reason and draw inferences and find answers that otherwise could not have been found.
For instance, there are ingredients that have other ingredients that are not mentioned on the label but the "other" ingredients can be inferred. Most of the time, Autolyzed Yeast Extract contains MSG, which is not mentioned directly on the label and this can be inferred with the use of ontology by having such specific relationships.
Broader Impact
Since a majority of the people use smartphones, applications, and most importantly the web, an awareness created via the web will have a higher impact as compared to other media. People often look up the ingredients that can have possible health risks and visit a number of websites before finalizing. Wouldn't it be helpful if the information could be consolidated and at your fingertips? Definitely, yes, and that is our goal. Even though this class project may be considered just to scratch the surface given the course time frame, we would like to continue working on this project and your feedback is always welcome.
Source and Copyright: http://www.takepart.com/foodinc/