Matchmaker site between biologist and ML experts

[hlapp]

From Scott Rifkin (UCSD):

One outcome that could be very useful would be something like a matchmaker site where a biologist, for example, could post a description of the data and the sort of analysis they have in mind, and ML experts could see if anything interests them to collaborate on. Or the converse. As a biologist I have some image data datasets and have in mind some datasets where ML could be a big help, but I don't really know of an efficient way to find a collaborator with the relevant expertise who might also find the particular problem interesting. And I imagine there are ML experts who don't know what datasets are floating around out there and who might have them that might pose exactly the sort of problem they are interested in. Some way to make it easier to find interested and relevant collaborators outside and beyond the few days of the workshop would be a very useful resource.

[DiamondKMG]

I really like this idea - you could have a "looking for" type tag system that could list not only the type of AI/ML you're trying to get at but also the level of help you need. Stack overflow q&a or full coauthor level collaboration for the data generators. Likewise a list of tags that might be useful for data/ computer scientists to find datasets that they find interesting.

[AnnaLew]

Great idea! I believe something similar is already in development (BioNet). I think it won't include posting the description of the data like suggested above, but rather the data owner will seek professionals to analyze it. Kind of like Fiverr for scientists. I might be wrong however, that's just how I understood the concept.

[hhsieh]

I like this idea and feel thrilled about it. Just one question - is the collaboration limited to Imageomics related or it can be broadened to include videos of insect interactions, crop drone imageries, or phenocams?

[cvstewart]

Glad you are excited. Collaboration is not at all limited to Imageomics!

…

On Wed, Aug 2, 2023 at 7:31 PM Hsun-Yi Hsieh ***@***.***> wrote: I like this idea and feel thrilled about it. Just one question - is the collaboration limited to Imageomics related or it can be broadened to include insect interactions, crop drone imageries, or phenocams? — Reply to this email directly, view it on GitHub <#1 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AD4SZBFXDDWPPDQRXG76TRDXTLPNVANCNFSM6AAAAAAY457WUE> . You are receiving this because you are subscribed to this thread.Message ID: ***@***.***>

[grenblau]

I imagined a general biology - machine learning matchmaking service. So any field.

[DavidCarlyn]

I like the idea. A general list of "open problems" for particular data would be a start. I think it would also be useful for ML scientists to includes lists of tools and their capabilities so the matchmaking is possible both ways. I feel like there should exists a platform for this, even just using wiki pages would be a start.

@AnnaLew Do you have a link to the BioNet you are referencing?

[adyork]

Late to the conversation here, but I'd be very interested to help identify potential datasets and promote a matchmaking service to folks who have image collections and want to collaborate!

[adyork]

I was able to talk to a few of you about this during the workshop and I'd love to keep some momentum going.

I'll second @DavidCarlyn

@AnnaLew Do you have a link to the BioNet you are referencing?

Looking at existing efforts before trying to recreate the wheel seems like a good idea!

From the biologist side of the matchmaking I'd love to pass on any information to biologists from ML experts about where they might discover possible datasets to work on, and if there are any standards data managers could put data in to make the collaboration easier. For example, getting datasets in huggingface, opinions on whether formatting images and annotiations using coco formats https://cocodataset.org/? Role of Kaggle competitions, etc.

[AnnaLew]

@adyork @DavidCarlyn sorry for the late response. The BioNet I was referring to is currently being developed by Marc Salit's team from MITRE. I don't think there's much about it online as it's still under development, but here they give an email that you can use to ask questions: https://www.mitre.org/events/synbiobeta-conference. From my understanding (I heard about it at a conference and talked with Marc about it), it will go public in around a year.

[adyork]

Thanks @AnnaLew I'll follow what's happing in that space.

@adyork @DavidCarlyn sorry for the late response. The BioNet I was referring to is currently being developed by Marc Salit's team from MITRE. I don't think there's much about it online as it's still under development, but here they give an email that you can use to ask questions: https://www.mitre.org/events/synbiobeta-conference. From my understanding (I heard about it at a conference and talked with Marc about it), it will go public in around a year.

[evodevosys]

[this is Scott Rifkin] I wasn't able to make the workshop in August, but wanted to follow up. Were there any ideas at the workshop about whether such a matchmaking service would be useful or how it could be implemented?

[adyork]

Anyone know of any matchmaker-type project or tools?

I looked a bit at MITRE's site and don't see anything that fits the description of what came up in this thread previously https://www.mitre.org/focus-areas/artificial-intelligence

Or any other suggestions I can pass along to scientists looking for ML dev connections? I've mentioned getting an imaging dataset in front of the huggingface community is an option. Which is more a "putting it out there and seeing what happens" approach. https://huggingface.co/docs/datasets/image_dataset

But for those looking for a closer connection or more formal collaboration, I don't have many suggestions other than offering to help make those connections where I can. But I mostly know marine science devs already saturated, and don't know many of a more purely ML folks that might be looking for unique challenges.