 So our final lightning round session for the day for big talk from small libraries is our international speaker. Very excited to have him with us today. Faizal Ariusintos is from Indonesia and is joining us from the Polytechnic Institute there. And he is going to be talking to us about and thank you so much for joining us so late in the day for you. Yes, of course. So he's going to talk about using technology to cope with the workload, which I think maybe we're trying to do a little bit today. I don't know. So go ahead and take it away and tell us all about it. All right. Hi everyone. I'm Faizal from Indonesia and today allow me to share my experience in using technology to copy the workload in our library in Polytechnic Institute of Nuclear Technology. And firstly, I want to introduce the background for our library. Our library is located on the second floor on this building, as you can see in the presentation. And we only have two staff here, including me and another is the head of library. When I joined the library, it was traditional rather than modern library. We need to write on paper when someone wants to borrow a book and students need to register themselves on a guest book when they enter the library. And unlike other university libraries, we don't have a repository, which is important for us. On the other hand, we have more than 1000 PDF files of final project from students. And it was a pandemic. And we went into lockdown the library was like a dead place. Since we kept this traditional way, we had no online service and no one came to the library. And so I decided to talk with the head of the library and said that we need to try another approach to make library live with some technology. And then he approved my proposal and creatively, our library got a better position and had many services for students. And one day, my friend shared information about the workshop held by the Ministry of Communication and Information Technology. It was part of Government Transformation Academy. I joined the workshop and learned about the science fundamentals. And we also learned about machine learning. And there were so much things that I can from this workshop. We learned from scratch the fundamentals and practicing in some real cases and have this passion to sharpen our understanding. This workshop aimed to allow participants to use this technology in their workplace. This is also the reason to answer why the participants were diverse public relations, IT staff, lecturers, and even librarians joined this program. And what is interesting is that it's new for me. And when there even been, I started thinking about adopting this technology in our library. So the first experience using this own library was coming from a student who tried to find a word, a machine name, on a final project documents. He couldn't find it on a title or abstract. So we need to dig deep into the full text, finding this word is possibly by open dbds file. And then we press control F. But it takes time. And if we assume one document can spend about half of minutes finding on who all documents can be done in almost nine hours without stop. So I created a process to filter certain text from bds file and it worked. We found three titles that contain this machine name. And after that we decided to use it as a part of our service. And in Indonesia, we adopt not nuclear subject taxonomy issued by the National Nuclear Energy Agency of Indonesia. This taxonomy spreads nuclear built into six subjects or competencies. And as I mentioned before, we have more than 1000 final project documents. And the good news is that we are going to have a repository. So we need to classify every document into one of the three subjects. And the problem is that we as librarians don't have a deep knowledge about nuclear energy. And sometimes we ask for help to the professor, lectures, our researchers to give us advice. But since the document are pretty large, we are afraid to ask for help because we also have jobs to do and classifying document is not that easy. So the workshop I mentioned about classifying documents and I tried to learn more about it. And after some tests, I finally finished the process that can help us with the giving advice for the proper subject of each document. In simple terms, we use documents already labeled from the National Nuclear Energy Agency repository and let this algorithm learn words. And the algorithm we use is KNN, learn and predict the right subject for our documents. And this process took about 30 minutes and we were happy because we could save time and concentrate on another test that need to be done. And today we have many services to give to our members. One of the newest is bibliometrics. The service helps students and our researchers to see trends or gaps on a specific subject to the keywords. And we usually use bibliographic data for taken from SCOPUS to make the network. We ask students to provide particular area of interest and then we search it on SCOPUS with scholar or else and or not the result. Using this data has a dilemma. Sometimes the keywords not clean enough to represent the real network. Librarians and librarians are the set concept and we have to often find this on other case like this. On the other hand, this database are the only source that we have. So we need to clean the data and find the root word on a keyword before we visualize it. And again, it take much time when we try to look one by one keywords that we have. Maybe it would be simpler if the data didn't have many keywords. When I learned about the science, I remember something useful. It's called word steaming, a process that helps us to find the root word. For example, words like librarians and librarian now merge into librarian, apples and apple as also one concept. Now it is apple. This brings us an innovation on our surface and create a better understanding of analysis, the network that occurred inside. We can save time to save this point. And at this point, I learned that we must be brave to try something new in a library with expanding our perspective, try to see from different angels that may benefit us and don't see technology as something that threaten us. Treat this as a tool that brings us a new level of satisfaction for the user and our library besides its benefit that I explained before. It's common thing that we as a librarian might have an adequate knowledge about new technologies. But I remember what Laosus said, journey of a thousand miles begin with a single step. We can start small to become big as long as we remain consistent. So that's all. If you have something to discuss or want to collaborate, use this link or scan the key airplane. Thank you. Thank you so much. Sorry, I'm just reading when to come in here. Thank you so much for all that. Your last comment there I think is very important. Don't be afraid to try. Some of that was some very intense, you know, machine learning is something that I'm sure a lot of librarians do not have experience or knowledge of. But definitely congratulations for going into doing that. I do have a comment from someone saying, having done consulting taxonomy work with specialized vocabulary and training machine learning in taxonomy starting from scratch as you did is daunting. Yes, I absolutely respect the effort this is taking in embracing the tools rather than fearing them. So for the librarian I think I use a application that called RapidMiner. It has a practical user interface and we don't need to learn about the code. We need to learn about the concept so we don't spend too much for coding. If you want to learn about it, you can contact me or find the website of RapidMiner or OrangeMiner. Absolutely. Yeah, there's so many tools we can use and we don't have to be a computer programmer. Yes. Awesome. Thank you so much. This is very helpful and interesting presentation. Thank you. Yes. All right. Yeah, thank you.