Featured research (5)
In this paper, we discuss the use of natural language processing (NLP) and artificial intelligence (AI) to analyse nutritional and sustainability aspects of recipes and food. We present the state of the art and some use cases, followed by a discussion of challenges. Our perspective on addressing these is that while they typically have a technical nature, they nevertheless require an interdisciplinary approach combining NLP and AI with expert domain knowledge to create practical tools and comprehensive analysis for the food domain.
Obesity and overweight in children and adolescents have reached 28% in the UK in 2016 (Health Survey for England, 2017). § The increasing number of childhood obesity might also have serious implications on the environment. § The food system currently contributes to about a quarter of global greenhouse gas emissions (GHGE), with generational and individual dietary choices influencing the magnitude of associated GHGE. § According to Notarnicola et al. (2017), animal protein products share a major part of the total environmental impacts of diets. § Little is known about the relationship between obesity, diets (in particular products high in animal protein) and climate change. § Aim: to identify the connection between the body mass index (BMI) of children under 10 years old and the GHGE of their dietary choices with a focus on GHGE from animal and non-animal protein sources. Methodology § We used the UK National Diet and Nutrition Survey (NDNS) database years 1-4 (2008/2009 – 2011/2012) to determine the number of obese and non-obese children (under 10 years of age) using the body mass index (BMI). § BMI average for boys and girls used were from WHO - between 16.8 and 19.3 were considered overweigh (n=511) and between 18.6 and 23 were considered obese (n=241). § We then estimated the GHGE associated with diets of these population group using data from Bates et al. (2019). § Then we calculated the average of GHGE per day (gCO2e/100g), the average of vegetable protein per GHGE (gCO2e/100g), the average of meat protein per GHGE (gCO2e/100g), the average of other protein sources per GHGE (gCO2e/100g). § We evaluated the relationship between obesity and GHGE to understand the impacts of unhealthy diets
Information related to greenhouse gas emissions (GHGE) embodied in the production and consumption of multiple foods (including meat and dairy) have become more available in recent years thanks to literature reviews and meta-analysis of life cycle assessment literature. However, there is limited matching of this information to dietary databases. This linkage is needed to investigate the climate change impacts of different dietary patterns, to formulate policies for helping to shift population’s eating habits towards healthy and sustainable diets. Linking a dietary database to GHGE is time consuming as well as effortful. These activities are typically not well documented which makes them hard to replicate. Furthermore, as each country has multiple dietary consumption and purchase surveys (which potentially are redesigned between each survey version), there is a potential for coding GHGE to global dietary databases resulting in months of labour. This is a major limitation to speeding up policy making promoting food sustainability, and making information related to global dietary sustainability widely available. Many global dietary databases are already harmonised to be comparable using the FoodEx2 system, a description and classification system developed and maintained by EFSA. It is currently used at global level with the support of FAO and WHO. FoodEx2 consists of a vocabulary of foods with assigned codes structured in a hierarchical manner, allowing the classification and description of foods reported in different types of data (e.g. food consumption, composition, or production method etc.). The linkage of GHGE databases to individual FoodEx2 codes would allow rapid matching to any previously harmonised FoodEx2 food datasets. This poster reports the results of a pilot study that mapped aggregated GHGE databases (Poore J. and Nemecek 2018) to the FoodEx2 classification. As GHGE databases of food products vary in scope of assessment and system boundary, this leads to differences in final GHGE values for each FoodEx2. These database differences in reported dietary GHGE are compared using different EU food consumption databases. The results suggest that this method allows us to provide rapid global GHGE related to diets comparing multiple GHGE data sources.
Food systems currently contribute about a quarter of global greenhouse gas emissions (GHGE), with generational and individual dietary choices influencing the magnitude of associated GHGE . The exhibit Take a Bite out of Climate Change addresses this issue developing outreach and educational materials to engage with the public in order to raise awareness about the impact of food choices, promote sustainable food consumption behaviours and empower consumers with accessible knowledge to make informed decisions. This poster summarises the impact of our outreach activities at the Royal Society Summer Science Exhibition (RSSSE) in London and Bluedot Festival (BDF) at Jodrell Bank Observatory in Maccleseld in July 2019. In general, both events received a varied audience, for instance, families, students, teachers, members of the public, soirees etc.