Foodways wants to develop a tool dealing with the topics of good food, sustainability, value chain transparency, future innovation… using machine learning and big data analysis methods. The tool should tap on massive amounts of unstructured data, which humans typically aren’t able to handle. By transforming and interpreting the data, the tool should be able to structure the information and generate relevant insights on selected questions.
the food industry's challenges
- How can sustainability professionals in the food industry keep up with an accelerating world?
- There’s too much information available on- and offline, and it becomes hard to decide what’s relevant. It’s often hard to take decisions, especially under pressure, without enough time, not knowing which data is relevant and/or should be taken into account.
- Companies, as their consumers, need more clear and relevant insights to take decisions.
the food industry's questions
very often the food industry poses itself questions for which she needs quick and relevant insights to make a decision. The following examples give you an idea of such questions:
- The sustainability leader of a corporation wants to understand the main sustainability challenges and risks the company will or may experience within the next years, based on the developments of the last months.
- The marketing department of a food SME is taking investment decisions for the newest product development. The department wants to understand the latest trends on the consumer food products sector and get an idea which mega-trends are gaining force.
- An entrepreneur wants to decide whether a vegan restaurant in the city centre will make sense and if the time is right in a specific location for a 100% vegan food offer. Also, he would like to know if a 100% value chain transparency is possible for his market, and if the consumers are demanding such solutions.
- A food corporation wants to know their role in the sustainability developments of the healthy and organic food developments, assessing the public's perception of its role, its competitors' role, as well as available technologies triggering more sustainable options. Furthermore, they specifically ask to search for other relevant unknown variables.
- A company willing to export its food products to a different market wants to have access to all needed informations to do so: food market regulations, main actors, consumer trends, among many others.
- A government official wants to have an overview on all facts and figures about food waste in his country and around the world: relevant data, most important reports, best practices, among other - all within 24h.
- A school wants to know what would be the best concept on sustainable and healthy food for its school, mainly based on best practices from schools with similar challenges and cultural realities as its own.
the tool's tasks
A technology based on the human-machine cooperation, where technology meets food experts.
- Information sources: After curating the information sources (human task), selecting them by quality and relevance, the tool should be able to work with the data crawled or accessed (machine).
- Possible IT tasks: The tool should use big-data analysis, machine learning and yield quick smart-data results (insights).
- Results: The tool will then deliver the relevant content to the food industry (private, public and not for profit organisations), in different formats. This is where the human role kicks in again, compiling this information into a report or just by sending the raw insights to the decision makers.
1. Information Sources
News articles / Blogs / RSS:
- Generalist (e.g. buzzfeed.com, guardian.com)
- Topic specific (e.g. sustainablefoodtrust.org, foodtank.com, ethicalfoods.com, civileats.com)
- Writers and journalists (e.g. michaelpollan.com)
- Food industry newsletters (e.g. Food Tech Connect; Food Dive; Café Future; Food Climate Research Network; Save Food)
agri-Food specific data
- National agri-food statistics databases (e.g. Agrarbericht.ch, Data.gov, USDA)
- Recipes databases (e.g. bbcgoodfood.com)
- Land Matrix
- Swiss Food Composition Database
- US Food Composition Database
- Open Food Facts
- Food Consumption in the US
- Global Hunger Data 2013 (IFPRI, Datahub.io)
- IFPRI Global Hunger Index (Interactive)
- Syngenta Open Data
- Facebook, LinkedIn, Youtube or Pinterest.
- Legislation (e.g. EU Food Legislation, Swiss Food Legislation, Portuguese Legislation, Portuguese Food Safety Legislation)
- European Food Information Resource
- Books, magazines or reports in PDF saved e.g. in a Dropbox-folder (or other) in the cloud
- Book examples: The Oxford Companion to Food; Food Rules: An Eater's Manual; Food Politics: How the Food Industry Influences Nutrition and Health
- Google Maps, Amazon Directory, Conferences notes...
2. Possible IT tasks:
- Crawling raw data
- Interpretation of possibly relevant data sets (e.g. finding correlations)
- Delivering insights and/or Visualisation of data sets
Via these possible methodologies & technologies:
- Storage and Index
- Web crawlers (e.g. Diffbot or open source web crawlers)
- Automatic Graphs
- Data analysis and statistics (e.g. correlations)
- Machine learning, artificial intelligence
- Everything is open and there's no limitation on the language or technology to be used
- Reports (PDF)
- Powerpoint Presentations (PDF)
- Search results (e.g. Google)
- Infographics (e.g. PDF, web-based)
- Sprea dsheets (e.g. Excel)