We are pleased to announce the publication of a paper validating Harmony on real-life data: Using natural language processing to facilitate the harmonisation of mental health questionnaires: a validation study using real-world data, authored by Eoin McElroy, Thomas Wood, Raymond Bond, Maurice Mulvenna, Mark Shevlin, George B. Ploubidis, Mauricio Scopel Hoffmann and Bettina Moltrecht, and published in BMC Psychiatry.
Our study aimed to evaluate the effectiveness of Natural Language Processing (NLP) in harmonising mental health questionnaires for cross-study research in areas such as mental health.
By comparing the semantic similarity of questionnaire items using NLP (the Sentence-BERT transformer model) with their actual correlation in a sample population, we found a moderate relationship (r = .48, p < .001) between the two measures. This suggests that NLP can accurately identify similar questions across different questionnaires.
While the NLP model showed promise in uncovering underlying patterns in the data, it required manual intervention to determine which relationships were truly relevant.
Our study showed that NLP can be a useful tool to match similar questions from different questionnaires, but it’s not perfect and should be used with caution.
A BibTeX entry for LaTeX users is
@article{mcelroy2024using,
title={Using natural language processing to facilitate the harmonisation of mental health questionnaires: a validation study using real-world data},
author={McElroy, Eoin and Wood, Thomas and Bond, Raymond and Mulvenna, Maurice and Shevlin, Mark and Ploubidis, George B and Hoffmann, Mauricio Scopel and Moltrecht, Bettina},
journal={BMC psychiatry},
volume={24},
number={1},
pages={530},
year={2024},
publisher={Springer}
}
How are research funders reacting to the AI governance vacuum? A recent article by Sense about Science, a leading independent charity that promotes the public interest in sound science and evidence, highlights the growing need for responsible AI governance in research. The article, titled Research funders tackle AI governance vacuum with pragmatic guidance, discusses the alarming gap between the rapid development and adoption of AI tools, and the lack of clear frameworks for their safe and ethical use.
Tech Talk at the AI|DL AI Meetup (London) Artificial Intelligence and Deep Learning for Enterprise Thomas Wood presents the Harmony project at the 19th AI and Deep Learning for Enterprise meetup on 8 October 2024. In case you missed the talk about Harmony on Tuesday at Civo Tech Junction with AI and Deep Learning for Enterprise sponsored by Daemon, you can now watch the recording of the live stream on AI|DL’s channel.