 Hi, my name is Paul Albert and on behalf of my colleague Terry Wheeler, I'd like to do an update on reciter, something that I've presented before at CNN. Okay. So, reciter is a powerful NIH funded publications management system. It's designed for academic medical centers or institutions. Here's how it works. We use institutional identity data to create a scholars identity. We retrieve candidate articles from PubMed. We use machine learning to estimate the likelihood a given scholar wrote a given article. And then we collect feedback from faculty and librarians. There's a set of web services built into reciter that syndicates the data with downstream systems. And then the biggest changes that have occurred since the last time I presented relate to the the web application where you can output reports and conduct bibliometric analyses. These are the components and ecosystem. So on the top there you have institutional identity systems and these are used to construct the profile. And then you can see into a reciter application, which is typically installed on AWS, but you can still locally as well. There's a separate service which integrates with PubMed. And it takes data from PubMed and then formats it in JSON and has some nice optimizations there. The reciter is displayed in the publication manager web interface data from eyesight, which is NIH is bibliometric service is fed into reciter as well. And then the data from reciter is exposed to downstream applications like our Vivo profile system department lab websites and our faculty review tool. So here's useful identity data about our people. So here's, here's one example, and all these data can be leveraged in order to make high quality article recommendations. Here at while Cornell we monitor output by 20,000 people, I think it's like 21,000 now, mainly academic appointees, especially full time faculty of which we have 2200 postdocs. We have our doctoral students publications, residents for our affiliate hospital, and then members of our clinical and translational science center. So now I am going to hop on over to reciter so I can share with you what the user experiences like. So Dr. Leiden. Okay, so here's how Dr. Leiden might curate his publications. There are suggested publications here, and then users can accept them and they go into the accepted queue. So here's a bunch of. So this is, this is the evidence score, a score of 10 indicates high confidence. And then if someone else like a librarian is trying to judge whether this article was written by Dr. Leiden, they can click on the show evidence and all the evidence is contained here. Okay, so that's kind of old hat. What's new though is the reporting interface. So, let's say that you're a departmental administrator, or anyone really, and you want to see publications that were authored by both doctors Leiden and Rafi. And then you only want last authored publications. You want to click search. And so we have 172 articles, and now I can export to our TF. So it's essentially a Word document. And then I go into Word. And then all the, all the names are bolded. And so we have a heuristic which matches people to author position. One thing. And you know it can actually print out like several hundred publications. At scale. So if you wanted, if you had like a couple hundred person identifiers, these are all members of our Cancer Center. So I'm going to grab these, these IDs and go back to find people. I'm going to paste all them in there. 214 people. I'm going to click on create reports. And this generates a report of all publications by all those individuals. I could further filter down by on, you know, author position or whatever. Or I can sort by like what are the most impactful publications according to NIH. All these data can be output in the context of a CSV or Excel file. So you can get an authorship report or an article report. So that's, that's useful. And then there's also the. I want to show you the bibliometric, the bibliometric report functionality as well. I'm going to go back to Dr. But search. So here's Dr. Wyden. Here h index is listed in each five index. I can click on generate bibliometric analysis. And then it's actually here and it's, it generates, you know, pretty, pretty instantly really instantly. So what we're doing here is we're outputting in a word documents. The most impactful publications with some groups by cases where Dr. Lyden's first or senior author position, or where he's any position. And then at the top here, this is a, this is a summary statement regarding how Dr. Wyden ranks relative to peers. So that's what this, that's what this NIH percentile is. And I won't spend time going into that. And then we also rank Dr. Lyden compared to peers that share the same academic rank. And at the bottom here, you can see that those that each index and then you can see an explanation of what all I just described. Another kind of nice functionality we've added is the ability to send emails, notification emails. So there are two types of notifications. One is when an article has been accepted on your behalf. So your faculty at someone else accepted one of your publications. And then the other case is where someone has. I'm sorry, when there are pending publications for you, so there's a suggestion, and you need to provide feedback on it. And this is what the notification looks like. We allow users to manage what that minimum evidence score for their notifications are so if they only want to get the high confidence notifications they can do so. Okay, so that that takes us through to that takes us through what I wanted to share with you. I will say this is for our users with this application is only available inside the network, our network at the moment, we have different roles. I want to give credit to all these individuals I want to also give credit to the people who funded us. Okay, so that's it. If you have any questions, feel free to follow up with me at my email listed there, check out the repository check out the paper. Thanks very much.