X Close

CBC Events

Home

Menu

Mike Kelly – science and evidence

By Carmen E Lefevre, on 5 September 2016

This project moves us to a new level of data assessment. Having come a long way in understanding the importance of evidence in policy decisions, it is now time to make evidence accessible and useable. This should close the gap between knowledge and action and supersede the notion of common sense.

HBCP is a revolution in the evidence base and its use. The applications are wide-spread and will include a platform for real ‘evidence-based policy’.

James Thomas – the user interface

By Carmen E Lefevre, on 5 September 2016

The interface will be straightforward to use and should enable users to access the knowledge generated in the project.

James will assess whether this system can be used to conducted automated systematic reviews and how good these are, compared to traditional human-based review efforts. He will also look at whether such a system can lead to evidence being utilised in novel ways and for novel purposes.

 

John Shawe-Taylor – Artificial Intelligence

By Carmen E Lefevre, on 5 September 2016

Can computers be programmed to show human levels of intelligence?

Thus far, despite some successes, like the DeepBlue chess-computer, AI has not been achieved. The sheer number of computations required to apply logic to complex problems is unfeasible. Machine learning on the other hand, is perhaps a more promising approach. It trains a system up to use probabilistic reasoning that can work out patterns in data. This approach may be a way forward for achieving artificial intelligence and has indeed shown success in playing games such as Go, which allow for exponentially more combinations than chess, rendering a brute-force logic approach unfeasible. As such, this approach will be taken in the HBCP to learn to populate the ontology.

Pol Mac Aonghusa – knowledge extraction

By Carmen E Lefevre, on 5 September 2016

“Isn’t this project just like a search engine?”

The behaviour change party: imagine you go to a party, where the guests double every hour, but you are only allowed to speak to 13 people, how easy would it be to keep track of all guests interests at the party? – Impossible. Pol uses this analogy to explain why the ontology is required, so that researchers can speak to more (or ideally all) people at the party.

Critically, the system to be developed is tasked to not only extract one or two relevant reports (like google perhaps would) but to extract ALL relevant reports and summarise them for easier reading. Moreover, it will need to weight studies and reports by their quality and importance without introducing bias or opinion.

The challenges of this project are 1) to ‘de-noise’ the data without losing information; 2) to understand what an author really meant; 3) join knowledge from many sources; 4) assign confidence to the results; and 5) make it work in a reasonable timeframe.

 

Marie Johnston – Collaboration between disciplines

By Carmen E Lefevre, on 5 September 2016

This project involves iterative interaction between three sciences: behavioural science, computer science, and information science.

It will start with behavioural scientists annotating existing literature. Next, computer scientists will aim to train a computer to annotate and extract information from the literature in the same way scientists do. Finally, information science will translate the extracted information into a query-able user interface for users of behavioural science output.

 

Robert West – the preliminary ontology

By Carmen E Lefevre, on 5 September 2016

This project is extremely challenging yet even a small step in the right direction will improve the way we currently work in the field of behavioural interventions.

Robert asks and unpacks the ‘the big question’ in behaviour change:

  1. Intervention and comparator: What is the behaviour change intervention and how is it delivered? Compared with what?
  2. Exposure: What is the reach of, and engagement with, the intervention?
  3. Population: Whose behaviour does the intervention aim to change?
  4. Setting: What is the setting in which the intervention is operating?
  5. Mechanism: How does the intervention work?
  6. Behaviour: What behaviour or behaviours is the intervention targeting?

In the first instance, this ‘big question’ will be tackled with a ‘use case’ of smoking cessation, an area with a huge amount of literature.

To develop the ontology, the project will heavily draw on Cochrane’s PICO Ontology and add additional fidelity were necessary.

Susan Michie’s introduction

By Carmen E Lefevre, on 5 September 2016

Prof. Susan Michie introduces the Human Behaviour Change project. It is funded by a collaborative award from the Wellcome Trust and involves collaboration between institutions, disciplines, and countries.

The vision is to revolutionise how we 1) build knowledge and understanding about behaviour change and 2) use this knowledge to answer real world questions: what works for what behaviour, for whom, in what setting and why?

The method will involve building an ontology of behaviour change interventions, an artificial intelligence system and a user interface that allows the system to be queried.

A range of problems currently exist for interpreting the behaviour change literature. More than 200 behaviour intervention evaluations are published per day, making it impossible to read them all. Moreover, these publications are not consistent in their use of terminology, and reported detail. Much research is not directly relevant the specific context of knowledge users, e.g. studies conducted in the US but applied in the UK. Finally, our current methods of reviewing the literature are too slow to react in a timely manner to advances in the field.

The current project aims to overcome these problems. Computers will be used to read the current literature and extract the relevant information and to reason and generate new hypotheses based on existing work.

 

 

Welcome

By cbc-events, on 5 September 2016

Welcome to the official blog of the Centre for Behaviour Change’s events.

From 5pm today we are hosting the Human Behaviour Change Kick-off Event. Follow live updates here.