The Food Sentiment Observatory: Exploiting New Forms of Data to Help Inform Policy on Food Safety & Food Crime Risks

Lead Research Organisation: University of Aberdeen
Department Name: Computing Science

Abstract

Social media and other forms of online content have enormous potential as a way to understand people's opinions and attitudes, and as a means to observe emerging phenomena - such as disease outbreaks. How might policy makers use such new forms of data to better assess existing policies and help formulate new ones?

This one year demonstrator project is a partnership between computer science academics at the University of Aberdeen and officers from Food Standards Scotland which aims to answer this question. Food Standards Scotland is the public-sector food body for Scotland created by the Food (Scotland) Act 2015. It regularly provides policy guidance to ministers in areas such as food hygiene monitoring and reporting, food-related health risks, and food fraud.

The project will develop a software tool (the Food Sentiment Observatory) that will be used to explore the role of data from sources such as Twitter, Facebook, and TripAdvisor in three policy areas selected by Food Standards Scotland:

- attitudes to the differing food hygiene information systems used in Scotland and the other UK nations;
- study of an historical E.coli outbreak associated with venison products to understand effectiveness of monitoring and decision making protocols;
- understanding the potential role of social media data in responding to new and emerging forms of food fraud.

The Observatory will integrate a number of existing software tools (developed in our recent research) to allow us to mine large volumes of data to identify important textual signals, extract opinions held by individuals or groups, and crucially, to document these data processing operations - to aid transparency of policy decision-making. Given the amount of noise appearing in user-generated online content (such as fake restaurant reviews) it is our intention to investigate methods to extract meaningful and reliable knowledge, to better support policy making.

Planned Impact

Who will benefit from this project?
The most obvious and immediate beneficiary of the proposed demonstrator project will be Food Standards Scotland. Other beneficiaries are:

- Scottish/UK Consumers
- Scottish Government Ministers
- Other Public Agencies (e.g. Food Standards Agency)
- Food and Drink Businesses
- Local Authorities
- Wider Publics
- Developers/Users of Social Media Analytics Software

How will they benefit from this project?
Food Standards Scotland aims to ensure that information and advice on food safety and standards, nutrition and labelling is independent, consistent, evidence-based and consumer-focused. Analysing new forms of data in real time or near real time has the potential to inform decision making and policy, through the provision of new strands of evidence. The project will have immediate impacts in terms of the specific policy use cases identified in this proposal; it will also have a longer-term impact by building data science and open policy making capacity in the organisation.

Scottish (and wider UK) consumers will be the ultimate beneficiaries of the work described in this proposal - as a result of FSS delivering a food and drink environment in Scotland that benefits, protects and is trusted by those consumers.

Scottish Government Ministers will benefit from the project results, as they support delivery of a number of the National Outcomes (see http://www.gov.scot/About/Performance/scotPerforms/outcome), specifically:

- 'Our lives are safe from crime, disorder and danger';
- 'We live longer, healthier lives';
- 'Our public services are high quality, continually improving, efficient and responsive to local people's needs'.

The demonstrator will introduce many FSS stakeholders, including food businesses and local authorities, to the potential of using new and emerging forms of data. It also will benefit other organisations with a similar remit, e.g. the Food Standards Agency.

Local authorities and police forces will benefit through the use of new and emerging forms of data to support development of food crime policy. Ultimately, this will improve the quality and volume of actionable intelligence in relation to such criminal activity - potentially leading to less public money being spent on covert surveillance to detect food fraud.

Developers of tools for social media analytics will benefit, through access to the opensource Observatory software and experience reports from the policy use cases.

Companies will benefit from the development of flexible and scalable analytical methods for contrastive opinion mining and summarisation, and context-aware data veracity analysis. Though developed for the food domain, we envisage our Observatory as a powerful analytics tool which is generic enough to be adapted and applied to a wide range of economic activities, such as business performance analysis and evaluating the effectiveness of marketing strategies.

Members of the public will be encouraged to think about the relationship between social media and other forms of online content, and policy development. They will also be encouraged to reflect on the veracity issues inherent in much of the online discourse relating to food.

Publications

10 25 50
 
Description The project investigated the potential of social media data to support policy makers working in the area of food safety, food crime and public health. Working closely with staff in Food Standards Scotland (FSS) the project team investigated three case study scenarios to develop better understanding of the issues and opportunities associated with use of these new forms of data. Scenarios were as follows: public awareness of the food hygiene information schemes in place in Scotland and England/Wales; understanding E.coli risk and illness, including historic incident management; intelligence gathering for food crime investigations. Each scenario led to the creation of a series of datasets (derived from Twitter and online newspaper articles) and supporting analysis/findings; at the same time the project team used an agile development methodology to develop a tool - The Food Sentiment Observatory - that is available as openSource software. Analysis of new forms of data in the policy context raises important questions around methodological soundness and transparency, and so the project also investigated these issues through case studies and a survey of FSS staff. An audit component was incorporated into the Observatory tool (built upon the earlier PRISM ontology) to document data capture, processing and analysis steps; this enables a report to be generated to contextualise the results of such analyses.
Exploitation Route The Observatory toolkit has the potential to be used by others to support their own social media data exploration activities. The ontology and associated software support for audit of social media data workflows could form the basis of a module within other tools to support transparent policy-making enabled by new forms of data.
Several Twitter datasets have been produced by the project; these may be of use to other researchers and developers of text analysis and text-mining software.
We are currently developing an experience report publication that will discuss each of the project case study scenarios, and will outline the challenges associated with identification of policy relevant questions, social media data access issues, as well as processing and analysis methods. This report will be of interest to a range of government data analysts and policy makers.
Sectors Agriculture, Food and Drink,Communities and Social Services/Policy,Environment,Government, Democracy and Justice

URL https://sites.google.com/view/foobs
 
Description The work of the project has already had direct impacts on our project partner Food Standards Scotland (FSS). Through our use of the Open Policy Making Toolkit to run project events, FSS have been introduced to these methods - and are now adopting them for their own policy discussions. Early results from the project demonstrated to FSS that their existing Food Hygiene Information Scheme has little or no recognition amongst social media users in Scotland (in contrast to clear recognition of the FSA Food Hygiene Rating System, used in England). Other work highlighted an emerging food safety issue relating to fruits such as Papaya, and showcased the potential for social media analytics to play a role in intelligence gathering for food-crime.
First Year Of Impact 2017
Sector Agriculture, Food and Drink
Impact Types Policy & public services

 
Description Introduction of Open Policy Making Toolkit to Food Standards Scotland
Geographic Reach Local/Municipal/Regional 
Policy Influence Type Influenced training of practitioners or researchers
Impact Through a series of policy-relevant scenarios, the project team introduced Food Standards Scotland (FSS) staff to the Open Policy Making Toolkit (PolicyLab, Cabinet Office). Approximately 20 FSS staff were exposed to a range of techniques and this subsequently led to adoption of the Toolkit by FSS.
 
Title A collection of Tweets used to explore causes of self-reported foodborne illnesses on social media 
Description Data collected from Twitter (twitter.com) social media platform (10 Nov 2017 - 18 Dec 2017) to explore causes of self-reported foodborne illnesses on social media from posts originating in Scotland, UK. The dataset contains Tweet IDs and keywords used to search for Tweets using a programatic access via the public Twitter API. In addition, this archive also includes keywords that were used to cluster retrieved Tweets into smaller groups of messages containing mentions of specific keywords. This includes lists of keywords describing ingredients, foods and drinks, cooking techniques, and domestic implements. Additional keywords relating to food and places associated with food (e.g. restaurants) were generated using an automated machine learning tool based on a set of seed keywords. Finally, the last set of keywords used to cluster retrieved Tweets includes a list of names of food businesses located in Glasgow, UK. 
Type Of Material Database/Collection of data 
Year Produced 2018 
Provided To Others? Yes  
Impact This data was used to shape the development and evaluation of the Food Sentiment Observatory platform developed in this project. Analysis based on these data were presented to policy makers from the Food Standards Scotland agency. The agency has reported a change of attitude in the context of their future plans regarding the use of social media data. 
URL http://reshare.ukdataservice.ac.uk/853375/
 
Title A collection of Tweets used to explore general food hygiene discourse and attitudes to the differing food hygiene information systems used in Scotland and the other UK nations 
Description Data collected from Twitter (twitter.com) social media platform (5 Aug 2017 - 28 Aug 2017) to explore general food hygiene discourse and attitudes to the differing food hygiene information systems used in Scotland and the other UK nations such as England and Wales (i.e. Food hygiene information scheme vs Food hygiene rating scheme). 
Type Of Material Database/Collection of data 
Year Produced 2018 
Provided To Others? Yes  
Impact This data was used to shape the development and evaluation of the Food Sentiment Observatory platform developed in this project. Analysis based on these data were presented to policy makers from the Food Standards Scotland agency. The agency has reported a change of attitude in the context of their future plans regarding the use of social media data. 
URL http://reshare.ukdataservice.ac.uk/853373/
 
Title A collection of Tweets used to explore the potential role of social media data in responding to new and emerging forms of food fraud 
Description Data collected from Twitter (twitter.com) social media platform (6 May 2018 - 16 May 2018) to explore the potential role of social media data in responding to new and emerging forms of food fraud reported on social media from posts originating in the UK. The dataset contains Tweet IDs and keywords used to search for Tweets using a programatic access via the public Twitter API. Keywords used in this search were generated using a machine learning tool and consisted of a combinations of keywords describing terms related to food and outrage. 
Type Of Material Database/Collection of data 
Year Produced 2018 
Provided To Others? Yes  
Impact This data was used to shape the development and evaluation of the Food Sentiment Observatory platform developed in this project. Analysis based on these data were presented to policy makers from the Food Standards Scotland agency. The agency has reported a change of attitude in the context of their future plans regarding the use of social media data. 
URL http://reshare.ukdataservice.ac.uk/853377/
 
Title A collection of Tweets used to study reports of food fraud related to fish products 
Description Data collected from Twitter (twitter.com) social media platform (8 June 2018 - 22 June 2018) to study reports of food fraud related to fish products on social media from posts originating in the UK. The dataset contains Tweet IDs and keywords used to search for Tweets using a programatic access via the public Twitter API. Keywords used in this search were generated using a machine learning tool and consisted of combinations of keywords describing terms related to fish and fake. 
Type Of Material Database/Collection of data 
Year Produced 2018 
Provided To Others? Yes  
Impact This data was used to shape the development and evaluation of the Food Sentiment Observatory platform developed in this project. Analysis based on these data were presented to policy makers from the Food Standards Scotland agency. The agency has reported a change of attitude in the context of their future plans regarding the use of social media data. 
URL http://reshare.ukdataservice.ac.uk/853378/
 
Title A collection of titles and publishers of articles used to explore historical reports of an E.coli incident 
Description Data collected from Lexis Nexis database (http://www.lexisnexis.co.uk/) including the result of a search for UK news articles mentioning phrase "errington cheese" for the period between July 2016 and Jan 2018. 
Type Of Material Database/Collection of data 
Year Produced 2018 
Provided To Others? Yes  
Impact This data was used to shape the development and evaluation of the Food Sentiment Observatory platform developed in this project. Analysis based on these data were presented to policy makers from the Food Standards Scotland agency. The agency has reported a change of attitude in the context of their future plans regarding the use of social media data. 
URL http://reshare.ukdataservice.ac.uk/853376/
 
Description Food Standards Scotland / FoObs 
Organisation Government of Scotland
Department Food Standards Agency (FSA), Scotland
Country United Kingdom 
Sector Public 
PI Contribution The research team have organised events with Food Standards Scotland staff to explore policy themes and have gathered/analysed a number of social media data sets to explore the potential of new forms of data in food safety, health and nutrition.
Collaborator Contribution Food Standards Scotland have provided access to staff (policy workshops, ad hoc meetings, data analysis sessions) and have been active participants in defining research questions and targeting data acquisition efforts. They have also been directly involved in analysis of findings from social media and casting these in a policy context.
Impact Too early to report items.
Start Year 2017
 
Title POPE Ontology 
Description Ontology (semantic knowledge base) used to support semantic annotation of provenance descriptions of the generation of policy evidence from social media data. 
Type Of Technology Software 
Year Produced 2018 
Open Source License? Yes  
Impact None to date 
URL https://foodsentimentobservatory.github.io/PoPE/index-en.html
 
Title The Food Sentiment Observatory Tool 
Description A collection of components of the Food Sentiment Observatory tool. The repository contains JAVA code for collecting Tweets and Python scripts for analysing collected data. A JAVA FX wrapper interface for a desktop version of the tool is also included in the repository. 
Type Of Technology Software 
Year Produced 2018 
Open Source License? Yes  
Impact None to date 
URL https://sites.google.com/view/foobs/the-observatory
 
Description Food & Social Media: Just #foodporn or a Force for Good? 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach Local
Primary Audience Public/other audiences
Results and Impact The event was organised as part of the programme for British Science Week 2018 in Aberdeen. Approximately 20 members of the public listened to short presentations from the academic PI (Edwards) and Food Standards Scotland representative (Pryde) describing their work using social media data to help shape food related policy in Scotland. The presentations were followed by a debate and discussion
Year(s) Of Engagement Activity 2018
 
Description Food & Social Media: More Useful Than You Might Think! 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach Regional
Primary Audience Public/other audiences
Results and Impact At a stall at the Huntly Hairst Farmer's Market during the Sunday of the Huntly Hairst, three members of the project team (David Corsar, Milan Markovic, Nikol Petrunova) ran a public engagement activity related to the use of social media data to improve food safety and standards. Attendees to the stall engaged in discussion on the topic, and were invited to take part in activities including voting on which social media platform they felt could provide the most useful data for improving food safety and standards, manually sorting and assessing social media data, and using a software tool to visualise and analyse a large social media data set.
Year(s) Of Engagement Activity 2017
 
Description Presentation to Scottish Government Data Analysts, Edinburgh, September 4 2018 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach Regional
Primary Audience Professional Practitioners
Results and Impact Approximately 30 analysts from across a range of Scottish Government departments attended a presentation about the Food Sentiment Observatory project given in Edinburgh, September 4 2018. There was an extensive discussion after the talk during which audience members expressed interest in the findings of the project, and sought clarification about the availability of social media data, and transparency of decision making based on large scale social media analytics.
Year(s) Of Engagement Activity 2018
 
Description Presentation to the PolicyLab team, Cabinet Office, London, 25 July 2018 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Policymakers/politicians
Results and Impact A presentation was given to members of the PolicyLab team at the Cabinet Office, London on 25 July 2018. The presentation described the Food Sentiment Observatory project and its use of the Open Policy Making (OPM) Toolkit as part of a larger social media analytics workflow. Following the presentation there was an extensive discussion with the PolicyLab team, during which they explored our experience of use of the OPM Toolkit and asked us to share use cases linking use of their toolkit to social media data gathering, analysis and any eventual policy findings.
Year(s) Of Engagement Activity 2018
 
Description Semantic modelling and utilisation of contextual information within¬†intelligent systems 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Postgraduate students
Results and Impact A guest lecture at the Xi'an Jiaotong-Liverpool University, China including an overview of several projects from our research group.
Year(s) Of Engagement Activity 2018
 
Description The Food Sentiment Observatory: Exploiting New Forms of Data to Help Inform Policy on Food Safety & Food Crime Risks 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach Local
Primary Audience Policymakers/politicians
Results and Impact A 40 minute presentation summarising project outcomes was given to an audience of approximately 60 employees of the Food Standards Scotland. The talk was followed by a discussion with the audience.
Year(s) Of Engagement Activity 2018
 
Description Workshop on New Forms of Data, Food & Food Policy, Aberdeen, 30 July 2018. 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Professional Practitioners
Results and Impact One day workshop on 'New Forms of Data, Food & Food Policy' held at the University of Aberdeen on July 30th, 2018. Speakers from the Food Sentiment Observatory project were joined by other academics and industry speakers. The audience comprised academics, local authority representatives, regional development professionals, and industry professionals. Participants reported increased knowledge and awareness of the potential of new forms of data in the food policy and food business sectors.
Year(s) Of Engagement Activity 2018
URL https://www.slideshare.net/abefoobs/presentations