Cadmore media player playing video Let's Prioritize Delivering the Content our Members Want, Not Just What We Have
Video Player
Accessibility
Using forms mode, you can utilize keyboard shortcuts to make usability more efficient.
Space bar will play and pause the video. Pressing escape will close all displays and pressing it again will jump to the end of the player.
You can press A to open the Keyboard Shortcut Dialog which also provides an audio description of the player and all of its features.
Language
1x
Play Speed
Adjust Volume
Transcript
Search
Result Count
of
Click to jump to result
Search
Re-center
Segments
Result Count
of
Click to jump to result
Search
Name:
Let's Prioritize Delivering the Content our Members Want, Not Just What We Have
Description:
Let's Prioritize Delivering the Content our Members Want, Not Just What We Have
Thumbnail URL:
https://cadmoremediastorage.blob.core.windows.net/b61fd969-92ef-47b7-89f6-f5e4a63bfea5/videoscrubberimages/Scrubber_1.jpg
Duration:
T00H30M16S
Embed URL:
https://stream.cadmore.media/player/b61fd969-92ef-47b7-89f6-f5e4a63bfea5
Content URL:
https://cadmoreoriginalmedia.blob.core.windows.net/b61fd969-92ef-47b7-89f6-f5e4a63bfea5/industry_breakout__molecular_connections_2024-05-29 (1080p).mp4?sv=2019-02-02&sr=c&sig=KykvTDCKW6XsFcH%2FwxJHzMF54l6E6GI1uHqyKnSZinU%3D&st=2025-05-24T18%3A05%3A40Z&se=2025-05-24T20%3A10%3A40Z&sp=r
Upload Date:
2024-12-03T00:00:00.0000000
Transcript:
Language: EN.
Segment:0 .
First of all, Thank you very much. Good afternoon to you all. I hope I'm visible. I always envy you, Stephen, on that. Just to quickly introduce molecular connections. We have been there for more than two decades servicing the publishing world.
But we don't exhibit we don't take a booth. So generally, many people don't know us. So just a quick glance at what we do. We have, I don't think, any conference. Cannot be complete without attending the reward. So we do a bit of both manual and automatic for publishers. When I say automatic is more based today, people talk about it being AI based.
It's largely machine learning. We do new products and workflow solutions. We do peer review services, finding peer reviewers, semantic tagging and discovery services and build platforms for publishers. So this particular product that I'm going to showcase, which we did for bone and joint publishing, is a new product, good revenue source for them, increase their member engagement by about 6,065% on their website is done by our SME team and this is the kind of team we have.
We are about 3,000 people with about 500 PhDs and 150 data scientists and about 300 software developers. So we work very high on the value chain. Good to see you, Rafael. And I did a poll with the SSP app on what do you think the content is needed. What kind of content on the content side, what do members want. And no surprise, accurate alert was among the top all related content.
Second and then accurate search was the third. And this just tells me that all our fraternity is quite well informed because we did a similar contextual inquiry with members of the bone and joint society and we got very similar results. And I'll just show you what we did. So this was about 18 months back. We did a market research of about $5,000 orthopedics, got about 1,200 responses, and 83% said.
Accurate search and updates on alerts was about 82% So it was nice to see that we are kind of on the same page here. The SSP fraternity. But my question is, what are we doing about it. We know our society members want accurate searches. We know that they want alerts. We know that they want all content on one place. So what are we doing about it was our question that we asked to the society.
These are some of the things that we noted that we received from the responses. So we need specialized domain searches. PubMed does not allow that. A society has maybe 5, 7, 10 journals. What about the others. We can't do specialized domain searches. We get a lot of irrelevancy into our search results. We don't have semantic level alert system.
Include non-journal content. That was a big thing. So people said, when I'm looking at journals, also give me standards, also give me videos, also give me every other thing that is connected to that particular let's say I'm searching for knee pain. I want everything on knee pain, the standards, also the videos also, and your own content and other content. So this was quite revealing and there are many other things that they told.
But this was a lot of it was this and therefore we came up with auto search. Auto search is an orthopedic knowledge discovery platform catering specifically to the orthopedic community. So we created we looked at all the content that is out there for orthopedics in the world, be it, standards, PubMed, abstracts, other journals or content podcasts, case reports, proceedings, you name it.
And we'll also search, which is now a repository of almost 400,000. Documents eight content types, including journal articles preprints proceedings, videos, standards. About 40% of it is open access. 60 is paid. This is a kind of a one stop portal for all that orthopedic needs, including alerts. It is SEO compliant.
So if you search for joint pain or knee pain on Google, it will be among the top three, top four. So it drives traffic back there. And this is the kind of content we brought together. This three lakh 35,000 odd journal articles and we update it every week. It's from 184 journals. So the editorial board at bone and joint publishing and the SME team in molecular connections in India sat down, decided these are the 180 odd journals that we are find relevant for orthopedics.
We went through all of them and then some journals. Of course, we covered back to back some journals. There was only one paper in Lancet which was important. There were two papers in PLOS, which was important. So we built a machine learning based relevancy engine to see what is relevant, what is not, and then again, mapped it all out. 10,000 preprints again, as I said, it's about every week that we update it.
3,500 podcasts, 500 techniques, 400 case reports, 400 standards, 2,700 videos. And I'll show you a demo of how it all comes together when you do semantic level searches. And how does it all come together. So before that, just let me show you what we did with. So how it all comes together is because we have built a very orthopedic specific ontology at the back end.
So how do orthopedics think based on that specialties were decided. So ankle and remember, this is also very useful for on the fly topic collections. So you want to make on the fly collections of oncology and orthopedic or spine injury or sports medicine. You can do that. Then these were further classified into root topics. This 15 specialties or 22 root topics and these 22 root topics had concepts about 20,000 concepts.
So this was at the back end of all the content. And this content was then used to tag everything the videos, the standards, the proceedings and everything, so that we can then have one search giving you all very accurate results. And these are all semantic level searches. So this slide is more meant to demonstrate the distribution of the concepts. So if you can see here, there are about 22.
It's not yet come, let me see. Here it is about 22,000 concepts. Anatomy has 2,500 and so on. And these concepts are the keywords, the metadata that connects everything. We keep these concepts store live. So we actually mine all the public ontologies that are out there, be it mesh, be it cabby, be it. In fact, we contribute to a lot of them.
So cab is done by us. We contribute to mesh a lot and we then take all these ontologies and keep them live so that all the keywords that we have generated are enriched with all the synonyms that are there, so that any kind of synonyms, search or semantic level search becomes easier. And all these at the back end is stored in a linked data manner. So they are all linked.
So a disease is linked to an anatomy. If they are known to have a link and let's say the body part is then linked to the broader anatomy part so that if you search in any which ways you can still get it semantically retrieved. So there is a kind of a knowledge graph behind all of this. And let me just show you an example of the knowledge graph. So let's say these are the metadata types, patients and analytical techniques, proteins, small molecules that we have extracted from an abstract or from a paper.
This metadata types. And this is a paper which says 20 year radiographic outcomes following single level. So now from the paper, the machine now knows that the patients are mentioned in a randomized trial, a procedure is mentioned, which is arthroplasty and a disease is mentioned, which is cervical disk collapse. And same way a second paper gives you a different metadata type, a different procedure, a different disease and a different drug.
And a third paper gives you a different thing. And this all are linked either through the broader I call you, I call the specialties, which is basic research. So small molecules are linked to patients and are linked to disease types and so forth, so that you can collect topic collections. And this is how orthopedic things. So they want to think about in terms of basic research, they want to think in terms of pharma research or they want to think in terms of clinical research, and that's how they do these searches.
So this whole knowledge graph is embedded behind the search and discovery interface. So this is my favorite slide because for any portal you need semantic relevance. So we have the metadata to improve the discoverability. Then this metadata is normalized and standardized and syntax is added. So that things talk to each other. A video talks to an abstract, an abstract talks to a standard and so forth.
And then we have an ontology behind to allow us content navigation. And then we have a linked data store on top of it at the backend. And so everything is connected. This is a simple slide saying, how do we generate relevancy? So we build a machine learning engine which said, as I said, I mean, for example, in Landsat, only 5.7% of the articles were orthopedic.
And we went through all the Landsat abstracts in the last three years, and this was the number we got 6% So we can't manually do this. We build a machine learning engine to do that. For public library of science, it was 0.14% which were relevant for injury, which is an Elsevier publication. Almost 26% was relevant for orthopedics. So we looked at all the content that is out there for the 184 journals that we wanted to be part of this portal and then ran our classifier to see what is relevant, what is not.
And of course, we did a human spot check, but we are fairly confident on our relevancy engine and we took all these into one portal. And as you saw in the previous slide, we had about 400,000 documents now. And then of course, we did a comparison about how it compares to PubMed. So because one of the things everybody talked about was we want accurate search.
And as PubMed like here, if you see DVT raises in lower limb, orthopedics was the search we did PubMed does not really understand lower limb as one word. It takes lower as one word and limb as one word and does give you a lot of things about lower, which has nothing to do with, limbs. But when the technology that I showed you earlier where we had a context, specific link graph built, it understands that lower limb is one part of the anatomy and DVT rates is talking about a disease.
And so therefore our searches were much more accurate than PubMed. The other problem statement is, of course, that we was articles preprints proceedings, broadcast videos, standards, everything in one go. And we did that in this portal. Another problem statement was contextual search. Now this is a classic example MRI as diagnostics before arthroscopic surgery MRI is very fairly commonly used, but PubMed or lots of portals that we tested on that does not really understand MRI as it just takes MRI as a string and then gives you whatever MRI is mentioned.
Whereas ortho search, the portal that we built understands MRI as a technique and therefore only gives you that particular results. So there is context built in again because of the link data graph. Again, this is very useful for alerts because.
Another example of muscle relaxant. When I'm talking about muscle relaxant, I want to get a medical condition, which is muscle relaxant and a medical condition against it. Very difficult to get this in PubMed or any other portal. Ortho search takes care of the different forms, synonyms, acronyms, variants and also follows conventions in which they are used in literature.
So for example, here is Achilles tendinopathy is extracorporeal shockwave therapy, Achilles tendinopathy, and it sees that this is the same and it will give you the same answers. I think I'll skip this, but I think you get the message. We have of course done comparisons of PubMed with autosearch and autosearch comes out way, way, way ahead in terms of stress factor is one word.
That would not get that. It takes dress differently and fracture differently and also shows you a lot more articles which have stress, but also search would not do that. Let me talk about the standards. Here is the usage. This is an interesting so we wanted to map the usage, what people are searching on the platform also so that it gives us some idea about what next to do in terms of our journal strategy as a publisher.
So this is a usage. This spike is because. Bone and joint presented the portal in a conference in November 23. But as you can see, the usage is really gradually increasing. The cumulative searches are also increasing quite a bit. The usage on the platform is very, very and this gives us a lot of confidence because this is our specialist thing and use so hip fracture arthroplasty and osteosarcoma are the highest search words.
They also come in with DUIs, so they just put the DUIs in and directly search through the DUIs. About 280 standards have already been searched. Metaarchive seems to be the preprint of their choice in terms of searching. And these are the authors who are most searched on the platform. It gives us a lot of insights like this same thing in terms of key phrases or specialties or title searches, and we map all this data.
So that's one part of it. What's next. And before I show you the demo of our search, which is a 2.5 minute demo, I just wanted to show you that next is topic collections. We may start that custom alerts premium service. We are seeing the usage on alerts very, very high. And to get an Roi on the portal, the publisher is thinking, can we charge for the alerts.
So we are experimenting with that. And of course, how can we not have an LLM conversational search on this. So that's the next step. This is already in beta. It would be launched in about most probably about four weeks. So the LLM looks something like this. So you can input your text. It says treatment options for compound ankle fractures in elderly with co-morbidities.
It's a pretty long statement. You put that in put the temperature that you want and you put the data set that you want. You can select here or to search or PubMed or whatever, and then you get the search results, which is something like this, which gives you a complete summary of a paper with the reference pulled in so you can check the references from where it has come so that it reduces the hallucinations that we are all aware of.
So with that, I would just show you a quick demo. I hope it works with Murphy's law. It never works, but let me try. OK, lovely. So this is so sorry. This doesn't work. I'm sorry, but it does not. Let me see if I can try this.
OK I'm sorry. It doesn't work here, but you can. Can you ask somebody in the corridor if somebody is there.
Yeah but. Any questions. Meanwhile, while I put up the demo on autosearch or what content why we went about this and. How we went about this. Are you talking about business model. Yes, sure. So the. So just to give you a cost perspective and then business model.
Yes so this video, I don't know why it is not coming. All right, so. We are. You're going to a separate window. Yes All right. Let's find that. So you are up here. One second.
Sorry so the. So there are three things that we looked at when we looked at from an Roi perspective. One was member engagement. So remember engagement score has improved by about 67% Two is. Money we are looking at advertisements. So conferences are not. They will have banners and we will show that. And advertising on that.
The third is that there will be a small amount. The fourth is we will build the community here. So there is an integration anymore where writing a paper and they can have their own. I think he had a citation here and then. I think the basic first idea was, can we at least give our members what they want and keep them without as a net.
Now, when we are seeing the success of the galaxy anterior labrum now in. Any other questions. Continued the question about the. One of the best sources for orthopedic information is actually the orthopedic device companies themselves who obviously want exposure to what.
Unlike pharma companies, you get somebody who adopts your hip striker, that's $1 billion win. So is that a potential for you to sponsored areas or how to videos or things like that. There is I assume there is this very nice section where we show the how to videos and. Devices because of the semantic level tagging. If the device is also mentioned on a paper, it can also pull that ad out.
And that's another. Making contact with somebody from the public. For publishing the. And so there are various but we are still in a very early stages here. As I said earlier, our plan was let's just make our members. And now this is not expensive. We did this in less than six months.
So the AI is now coming. Then we go out and play the video. Yes, please. All right. Thank you. Get back. Yep, absolutely. And I'll stick in here with you to get you switched back over, OK.
Yes All right. What should. So we weren't set up for laptop audio. So that's the problem. Is the laptops not connected to the mixer because that wasn't a part of the setup. No problem.
The is not missing. They are not connected. But I mean, you will see that it is a very, very clean interface. The topics on the left hand side, you can select by journals, you can select by and on the impact factor is mentioned. All the results are integrated with standards and videos and everything else. These are the kind of things.
Sorry, the voices are not good, but. But that's what it is. And then you can put alerts. There are related topics. Here is where the space is for the conferences where we are going to advertise conferences. This keywords are the keywords by which orthopedic search and these are very good metadata for the publisher.
So one of the things that we are doing is to take these keywords and put it back into a typhoon where this is posted. So that the search and discovery can also improve substantially. This is semantic level alerts. You get scheduler alerts seven days, 14 days, 10 days. We've got fantastic breakthroughs from the back to back to. You can also do an alert otherwise institution wise, keyword wise.
You can, of course, bookmark your favorites. You can. It's up on the net. The search darkness. You got the look and. Why are you doing this for.
One more society. Very similar approach. We already have the technology. We just have the friendship for another society in medieval times. And that was just to be up and running about. These are specialties because of speciality. Again, SEO becomes very easy and.
Then he has sexuality pages on your website. And again, this is compliant with your schema or compliant having videos and research. I could go on, but I think you get. Thanks do you think. An engineering. So you said it's primarily for the membership, but it's open to all.
If you want to implement more coming in. But it is exhibit for the members to go. It was a little small. How are you distinguishing on your own. What is from journal and what is. What is a podcast. Because they all look similar.
Yeah I wish I had the thing. Yeah, please if you can. Absolutely there is a separate tab or three print and you can just click the free print and it will only show you the free trade. But you are also here. You will also have a tab, which will show you what it is. So right now these are all abstracts, but if you see this tab here, which will show that it's free.
Plus there is a tab up, which you can only filter through print or video. What about when people hit a paywall? They're going to another article, another journal, and then they hit a paywall. That's a good question. So we will not unless it is open access, we will not have the text.
We will stop at abstract. So we'll skip the abstract and that's. Yeah and then we can buy it on the road at that point. Sorry, ma'am. You had a question. There are no more questions. I'm sorry. So in mixing formats with search.
It's often best to do your indexing full text of an article, but only an abstract of this or just a transcript of. How do you normalize relevance across wildly different. Well, that's a great question. It took us a lot of time to do that because it is a collaborative effort where we kind of want to sit with the editorial team and they tell us this is what they feel is more relevant and not.
And then we tweak our engine based on that. So for videos, it's a different engine in terms of weightages and transcripts, it's very different in podcasts, but for abstracts it's different. For text, it is different. Again, we have limits on full text. We are. Not having more than 20 keywords, Max. So then we score the keywords and then have it in abstract.
We have 10 keywords and so forth. So OK. Thank you so much for coming. I really appreciate your time.
Get out of there. You can watch me do it. We know that your proprietary information isn't there. So are you going to walk back there. So we love with them. So they want and then to be included in your index.
Are you.
Segment:0 .
First of all, Thank you very much. Good afternoon to you all. I hope I'm visible. I always envy you, Stephen, on that. Just to quickly introduce molecular connections. We have been there for more than two decades servicing the publishing world.
But we don't exhibit we don't take a booth. So generally, many people don't know us. So just a quick glance at what we do. We have, I don't think, any conference. Cannot be complete without attending the reward. So we do a bit of both manual and automatic for publishers. When I say automatic is more based today, people talk about it being AI based.
It's largely machine learning. We do new products and workflow solutions. We do peer review services, finding peer reviewers, semantic tagging and discovery services and build platforms for publishers. So this particular product that I'm going to showcase, which we did for bone and joint publishing, is a new product, good revenue source for them, increase their member engagement by about 6,065% on their website is done by our SME team and this is the kind of team we have.
We are about 3,000 people with about 500 PhDs and 150 data scientists and about 300 software developers. So we work very high on the value chain. Good to see you, Rafael. And I did a poll with the SSP app on what do you think the content is needed. What kind of content on the content side, what do members want. And no surprise, accurate alert was among the top all related content.
Second and then accurate search was the third. And this just tells me that all our fraternity is quite well informed because we did a similar contextual inquiry with members of the bone and joint society and we got very similar results. And I'll just show you what we did. So this was about 18 months back. We did a market research of about $5,000 orthopedics, got about 1,200 responses, and 83% said.
Accurate search and updates on alerts was about 82% So it was nice to see that we are kind of on the same page here. The SSP fraternity. But my question is, what are we doing about it. We know our society members want accurate searches. We know that they want alerts. We know that they want all content on one place. So what are we doing about it was our question that we asked to the society.
These are some of the things that we noted that we received from the responses. So we need specialized domain searches. PubMed does not allow that. A society has maybe 5, 7, 10 journals. What about the others. We can't do specialized domain searches. We get a lot of irrelevancy into our search results. We don't have semantic level alert system.
Include non-journal content. That was a big thing. So people said, when I'm looking at journals, also give me standards, also give me videos, also give me every other thing that is connected to that particular let's say I'm searching for knee pain. I want everything on knee pain, the standards, also the videos also, and your own content and other content. So this was quite revealing and there are many other things that they told.
But this was a lot of it was this and therefore we came up with auto search. Auto search is an orthopedic knowledge discovery platform catering specifically to the orthopedic community. So we created we looked at all the content that is out there for orthopedics in the world, be it, standards, PubMed, abstracts, other journals or content podcasts, case reports, proceedings, you name it.
And we'll also search, which is now a repository of almost 400,000. Documents eight content types, including journal articles preprints proceedings, videos, standards. About 40% of it is open access. 60 is paid. This is a kind of a one stop portal for all that orthopedic needs, including alerts. It is SEO compliant.
So if you search for joint pain or knee pain on Google, it will be among the top three, top four. So it drives traffic back there. And this is the kind of content we brought together. This three lakh 35,000 odd journal articles and we update it every week. It's from 184 journals. So the editorial board at bone and joint publishing and the SME team in molecular connections in India sat down, decided these are the 180 odd journals that we are find relevant for orthopedics.
We went through all of them and then some journals. Of course, we covered back to back some journals. There was only one paper in Lancet which was important. There were two papers in PLOS, which was important. So we built a machine learning based relevancy engine to see what is relevant, what is not, and then again, mapped it all out. 10,000 preprints again, as I said, it's about every week that we update it.
3,500 podcasts, 500 techniques, 400 case reports, 400 standards, 2,700 videos. And I'll show you a demo of how it all comes together when you do semantic level searches. And how does it all come together. So before that, just let me show you what we did with. So how it all comes together is because we have built a very orthopedic specific ontology at the back end.
So how do orthopedics think based on that specialties were decided. So ankle and remember, this is also very useful for on the fly topic collections. So you want to make on the fly collections of oncology and orthopedic or spine injury or sports medicine. You can do that. Then these were further classified into root topics. This 15 specialties or 22 root topics and these 22 root topics had concepts about 20,000 concepts.
So this was at the back end of all the content. And this content was then used to tag everything the videos, the standards, the proceedings and everything, so that we can then have one search giving you all very accurate results. And these are all semantic level searches. So this slide is more meant to demonstrate the distribution of the concepts. So if you can see here, there are about 22.
It's not yet come, let me see. Here it is about 22,000 concepts. Anatomy has 2,500 and so on. And these concepts are the keywords, the metadata that connects everything. We keep these concepts store live. So we actually mine all the public ontologies that are out there, be it mesh, be it cabby, be it. In fact, we contribute to a lot of them.
So cab is done by us. We contribute to mesh a lot and we then take all these ontologies and keep them live so that all the keywords that we have generated are enriched with all the synonyms that are there, so that any kind of synonyms, search or semantic level search becomes easier. And all these at the back end is stored in a linked data manner. So they are all linked.
So a disease is linked to an anatomy. If they are known to have a link and let's say the body part is then linked to the broader anatomy part so that if you search in any which ways you can still get it semantically retrieved. So there is a kind of a knowledge graph behind all of this. And let me just show you an example of the knowledge graph. So let's say these are the metadata types, patients and analytical techniques, proteins, small molecules that we have extracted from an abstract or from a paper.
This metadata types. And this is a paper which says 20 year radiographic outcomes following single level. So now from the paper, the machine now knows that the patients are mentioned in a randomized trial, a procedure is mentioned, which is arthroplasty and a disease is mentioned, which is cervical disk collapse. And same way a second paper gives you a different metadata type, a different procedure, a different disease and a different drug.
And a third paper gives you a different thing. And this all are linked either through the broader I call you, I call the specialties, which is basic research. So small molecules are linked to patients and are linked to disease types and so forth, so that you can collect topic collections. And this is how orthopedic things. So they want to think about in terms of basic research, they want to think in terms of pharma research or they want to think in terms of clinical research, and that's how they do these searches.
So this whole knowledge graph is embedded behind the search and discovery interface. So this is my favorite slide because for any portal you need semantic relevance. So we have the metadata to improve the discoverability. Then this metadata is normalized and standardized and syntax is added. So that things talk to each other. A video talks to an abstract, an abstract talks to a standard and so forth.
And then we have an ontology behind to allow us content navigation. And then we have a linked data store on top of it at the backend. And so everything is connected. This is a simple slide saying, how do we generate relevancy? So we build a machine learning engine which said, as I said, I mean, for example, in Landsat, only 5.7% of the articles were orthopedic.
And we went through all the Landsat abstracts in the last three years, and this was the number we got 6% So we can't manually do this. We build a machine learning engine to do that. For public library of science, it was 0.14% which were relevant for injury, which is an Elsevier publication. Almost 26% was relevant for orthopedics. So we looked at all the content that is out there for the 184 journals that we wanted to be part of this portal and then ran our classifier to see what is relevant, what is not.
And of course, we did a human spot check, but we are fairly confident on our relevancy engine and we took all these into one portal. And as you saw in the previous slide, we had about 400,000 documents now. And then of course, we did a comparison about how it compares to PubMed. So because one of the things everybody talked about was we want accurate search.
And as PubMed like here, if you see DVT raises in lower limb, orthopedics was the search we did PubMed does not really understand lower limb as one word. It takes lower as one word and limb as one word and does give you a lot of things about lower, which has nothing to do with, limbs. But when the technology that I showed you earlier where we had a context, specific link graph built, it understands that lower limb is one part of the anatomy and DVT rates is talking about a disease.
And so therefore our searches were much more accurate than PubMed. The other problem statement is, of course, that we was articles preprints proceedings, broadcast videos, standards, everything in one go. And we did that in this portal. Another problem statement was contextual search. Now this is a classic example MRI as diagnostics before arthroscopic surgery MRI is very fairly commonly used, but PubMed or lots of portals that we tested on that does not really understand MRI as it just takes MRI as a string and then gives you whatever MRI is mentioned.
Whereas ortho search, the portal that we built understands MRI as a technique and therefore only gives you that particular results. So there is context built in again because of the link data graph. Again, this is very useful for alerts because.
Another example of muscle relaxant. When I'm talking about muscle relaxant, I want to get a medical condition, which is muscle relaxant and a medical condition against it. Very difficult to get this in PubMed or any other portal. Ortho search takes care of the different forms, synonyms, acronyms, variants and also follows conventions in which they are used in literature.
So for example, here is Achilles tendinopathy is extracorporeal shockwave therapy, Achilles tendinopathy, and it sees that this is the same and it will give you the same answers. I think I'll skip this, but I think you get the message. We have of course done comparisons of PubMed with autosearch and autosearch comes out way, way, way ahead in terms of stress factor is one word.
That would not get that. It takes dress differently and fracture differently and also shows you a lot more articles which have stress, but also search would not do that. Let me talk about the standards. Here is the usage. This is an interesting so we wanted to map the usage, what people are searching on the platform also so that it gives us some idea about what next to do in terms of our journal strategy as a publisher.
So this is a usage. This spike is because. Bone and joint presented the portal in a conference in November 23. But as you can see, the usage is really gradually increasing. The cumulative searches are also increasing quite a bit. The usage on the platform is very, very and this gives us a lot of confidence because this is our specialist thing and use so hip fracture arthroplasty and osteosarcoma are the highest search words.
They also come in with DUIs, so they just put the DUIs in and directly search through the DUIs. About 280 standards have already been searched. Metaarchive seems to be the preprint of their choice in terms of searching. And these are the authors who are most searched on the platform. It gives us a lot of insights like this same thing in terms of key phrases or specialties or title searches, and we map all this data.
So that's one part of it. What's next. And before I show you the demo of our search, which is a 2.5 minute demo, I just wanted to show you that next is topic collections. We may start that custom alerts premium service. We are seeing the usage on alerts very, very high. And to get an Roi on the portal, the publisher is thinking, can we charge for the alerts.
So we are experimenting with that. And of course, how can we not have an LLM conversational search on this. So that's the next step. This is already in beta. It would be launched in about most probably about four weeks. So the LLM looks something like this. So you can input your text. It says treatment options for compound ankle fractures in elderly with co-morbidities.
It's a pretty long statement. You put that in put the temperature that you want and you put the data set that you want. You can select here or to search or PubMed or whatever, and then you get the search results, which is something like this, which gives you a complete summary of a paper with the reference pulled in so you can check the references from where it has come so that it reduces the hallucinations that we are all aware of.
So with that, I would just show you a quick demo. I hope it works with Murphy's law. It never works, but let me try. OK, lovely. So this is so sorry. This doesn't work. I'm sorry, but it does not. Let me see if I can try this.
OK I'm sorry. It doesn't work here, but you can. Can you ask somebody in the corridor if somebody is there.
Yeah but. Any questions. Meanwhile, while I put up the demo on autosearch or what content why we went about this and. How we went about this. Are you talking about business model. Yes, sure. So the. So just to give you a cost perspective and then business model.
Yes so this video, I don't know why it is not coming. All right, so. We are. You're going to a separate window. Yes All right. Let's find that. So you are up here. One second.
Sorry so the. So there are three things that we looked at when we looked at from an Roi perspective. One was member engagement. So remember engagement score has improved by about 67% Two is. Money we are looking at advertisements. So conferences are not. They will have banners and we will show that. And advertising on that.
The third is that there will be a small amount. The fourth is we will build the community here. So there is an integration anymore where writing a paper and they can have their own. I think he had a citation here and then. I think the basic first idea was, can we at least give our members what they want and keep them without as a net.
Now, when we are seeing the success of the galaxy anterior labrum now in. Any other questions. Continued the question about the. One of the best sources for orthopedic information is actually the orthopedic device companies themselves who obviously want exposure to what.
Unlike pharma companies, you get somebody who adopts your hip striker, that's $1 billion win. So is that a potential for you to sponsored areas or how to videos or things like that. There is I assume there is this very nice section where we show the how to videos and. Devices because of the semantic level tagging. If the device is also mentioned on a paper, it can also pull that ad out.
And that's another. Making contact with somebody from the public. For publishing the. And so there are various but we are still in a very early stages here. As I said earlier, our plan was let's just make our members. And now this is not expensive. We did this in less than six months.
So the AI is now coming. Then we go out and play the video. Yes, please. All right. Thank you. Get back. Yep, absolutely. And I'll stick in here with you to get you switched back over, OK.
Yes All right. What should. So we weren't set up for laptop audio. So that's the problem. Is the laptops not connected to the mixer because that wasn't a part of the setup. No problem.
The is not missing. They are not connected. But I mean, you will see that it is a very, very clean interface. The topics on the left hand side, you can select by journals, you can select by and on the impact factor is mentioned. All the results are integrated with standards and videos and everything else. These are the kind of things.
Sorry, the voices are not good, but. But that's what it is. And then you can put alerts. There are related topics. Here is where the space is for the conferences where we are going to advertise conferences. This keywords are the keywords by which orthopedic search and these are very good metadata for the publisher.
So one of the things that we are doing is to take these keywords and put it back into a typhoon where this is posted. So that the search and discovery can also improve substantially. This is semantic level alerts. You get scheduler alerts seven days, 14 days, 10 days. We've got fantastic breakthroughs from the back to back to. You can also do an alert otherwise institution wise, keyword wise.
You can, of course, bookmark your favorites. You can. It's up on the net. The search darkness. You got the look and. Why are you doing this for.
One more society. Very similar approach. We already have the technology. We just have the friendship for another society in medieval times. And that was just to be up and running about. These are specialties because of speciality. Again, SEO becomes very easy and.
Then he has sexuality pages on your website. And again, this is compliant with your schema or compliant having videos and research. I could go on, but I think you get. Thanks do you think. An engineering. So you said it's primarily for the membership, but it's open to all.
If you want to implement more coming in. But it is exhibit for the members to go. It was a little small. How are you distinguishing on your own. What is from journal and what is. What is a podcast. Because they all look similar.
Yeah I wish I had the thing. Yeah, please if you can. Absolutely there is a separate tab or three print and you can just click the free print and it will only show you the free trade. But you are also here. You will also have a tab, which will show you what it is. So right now these are all abstracts, but if you see this tab here, which will show that it's free.
Plus there is a tab up, which you can only filter through print or video. What about when people hit a paywall? They're going to another article, another journal, and then they hit a paywall. That's a good question. So we will not unless it is open access, we will not have the text.
We will stop at abstract. So we'll skip the abstract and that's. Yeah and then we can buy it on the road at that point. Sorry, ma'am. You had a question. There are no more questions. I'm sorry. So in mixing formats with search.
It's often best to do your indexing full text of an article, but only an abstract of this or just a transcript of. How do you normalize relevance across wildly different. Well, that's a great question. It took us a lot of time to do that because it is a collaborative effort where we kind of want to sit with the editorial team and they tell us this is what they feel is more relevant and not.
And then we tweak our engine based on that. So for videos, it's a different engine in terms of weightages and transcripts, it's very different in podcasts, but for abstracts it's different. For text, it is different. Again, we have limits on full text. We are. Not having more than 20 keywords, Max. So then we score the keywords and then have it in abstract.
We have 10 keywords and so forth. So OK. Thank you so much for coming. I really appreciate your time.
Get out of there. You can watch me do it. We know that your proprietary information isn't there. So are you going to walk back there. So we love with them. So they want and then to be included in your index.
Are you.
Form Title Download
Transcript
Form Title Bookmark
Entire Media
Segment(s)
Custom Clip
Start At:
End At:
Form Title Share
Entire Media
Segment(s)
Custom Clip
Start At:
End At:
Form Title Accessibility and Keyboard Shortcuts
The cadmore media player is a professional video and audio player experience designed to deliver media as well as relevant transcript, segmentation, and metadata. On the screen is a traditional html video player with buttons such as play, pause, forward, captioning, language and play speed selection and more. Adjacent to the video player are generally tabs, one of which shows a scrolling transcript of the video and the other shows any segmentation of the media, sometimes referred to as chapters. At the top of the video is a menu bar containing an Explore button and Tools button. The Explore button allows you to dig deeper into the video by opening up an explore dialog which contains tabs for a visual navigation of the video using thumbnails, an about tab which contains rich metadata about the media, a segments tab which contains further information about the video, and a related tab which contains links to other media related to this media. The tools button opens a menu with options for sharing the media on social media, creating bookmarking links, creating embed codes for your website, generating citations, and more.
Keyboard Shortcuts
Keyboard shortcuts only work while you are in forms mode. By entering forms mode, you will be able to quickly access portions of the media player with single key commands.
Spacebar: Play/Pause Toggle
I: Info Menu,
O: Tools Menu
T: Transcript Tab,
S: Segments Tab
F: Forward 15,
R: Rewind 15
L: Languages,
P: Playspeed
M: Mute Toggle,
V: Volume Bar
N: Video Only Toggle,
B: Full Screen Toggle
C: Captions Toggle
Escape: Closes dialogs and menus / Exits Player
Z: Next Transcript Chapter
X: Previous Transcript Chapter
Q: Next Transcript Paragraph
W: Previous Transcript Paragraph
Entire Media
Segment(s)
Custom Clip
Start At:
End At:
Embed Size
-
You Size - Responsive
-
We Size - Responsive
-
Picture Overlay Popup
-
400 x 225
-
512 x 288
-
400 x 700
-
768 x 1024
-
1024 x 512
-
1280 x 608
Form Title Cite
No citation provided.
Info
Segments
Related
Thumbnails
Exit Clip Mode
You have requested to see a portion of the video outside of the clip.
Click OK to switch to the full video or click if you would like to stay inside the clip. You can always go back to the clip with Tools and Clip Mode