 yma ynghylch, ac yn ymddangos am yr AOC, am y cwm iawn. Felly, rydych chi'n gweithio'n cyfrannu cyfrannu wedi'i gweithio'r cwm iawn. Rydych chi'n gweithio'r cwm iawn, yn ei gydag, yn ymddangos i'r unig i April. Felly, mae'n gweld y ffyrdd sy'n amser i amser i Anzin, am Peter ac Naveen, mae'n dweud i'n ddigon i'r gweithio i fynd i'n gweithio. Ond yn ymgyrch, sydd ydych yn gweithio'r gwaith, ond mae'r grwp ffasilydd ffordd yw'r gweithio. Yn y ffordd, mae'n rhan o'r link yn y chat, yn y moment, ifrwyd, mae'n gweithio'r gweithio'r gweithio. Felly, fydd ddim yn fwy o unig gael, dwi'n fwy o enw, ac mae'n gweinio i'r hunain. A mae'n gweinio o'r informeith anodd o'r hyn sy'n gweinio'r hyn arno gyda'r hynny. A mae'n gweinio i'r hyn ac mae'n gweinio i'ch gweinio i'r hynny o'r pwysigau i'r blog ac oedd gan hyn o'r pwysigau. in there. So this is the first of a small series. The only bit of advance notice I'll give you is one that is about to come about, which is Dr Helen Crompton. Now she's someone I've known for about 10 years. So I've managed to persuade her to do a chat GTP AI roll up or a summary of all of our webinar series on this, and she'll be doing that on April the 8th. So it's a bit of a, and we got it for free. So that was the important thing. And obviously I've just done the advance notice now so that if you're aware of it, you know you can look her up and so on. But that's a but we will have others in the next couple of months. So watch out for the notices. So I will end that bit and I shall now go in and move into the main presentation for today. And I will pass in to Yasmin. Yasmin. Okay. Over to you. Yep, yes, thank you. Thank you very much introduction. Okay, so thank you very much for the special interesting group provided this opportunity for us. So and we are very delighted to have this opportunity to share our some experience and with all of you. Just mentioned, this is a collaborative work with our colleagues Peter Nico and Navid Kehan. So what I'm going to talk about here, initially I want to cover three major aspects. One is the talk about the student engagement. So what we we can use to try to capture understand the student engagement. And the second part related to challenges from the chat GPT from teaching on the learning perspectives. And the third part is a share some experience regarding how to incorporate the machine learning into teaching activities. But unfortunately, so on the way, perhaps don't have much of time to talk about the third part and what I can do. So and I already send the link to the chat box. So and if you are interested and you can click that link to see something shortly. Okay. So we are from Ulster University. University, Ulster University basically is multiple campus university across different regions. And we have Bairfast campus. This one is based on the Bairfast. We have Jordan's Town in Jordan's Town and the McGee in McGee Town and a city, Dairis City. And also we have the headquarter of the university in Corrine campus. So university has a strategy and called the people people place and the partnership basically try to unlock all of the potentials from the three major aspect, build the sustainable future for all of the society's things. So this is our the university's strategy. So everything the teaching and learning research is this strategy basically provides the role model for us conducted teaching, learning and the research. It is a big umbrella. So where we are, so all of us based on Bairfast campus just mentioned now you can click the link to see where the Bairfast campus it is and what it looks like. I give you a few seconds. Okay. So can you click that link and to see it. Okay. So this is a new campus and we just moved in so from the last year, from last year so and from this the digital and the view and you can look at the different and the part of the campus from Main Street even to the inside of the building. Okay. So perhaps if you are interested and you can view it later. Okay. So let me to move on the next part. So the first I like to see a little bit about the impact emerging technologies and how this emerging technology affect and the learning behavior. So all the past years and high education actually helped be impact by various challenges. Clearly is a pandemic. So a pandemic disruption and also include some government policy change. I think the important part here is some emerging technology actually the key driver to lead the we call the digital transformation. So this kind of transmission clearly to see for example during pandemic period and we put all of the things into the online. So and from the traditional class teaching into the online teaching. So since then there's a large body of research work talk about the digital pedagogies and talk about the digital curriculum design did even talk about the digital universities. So and finally Payson in 2020 first published one article. So and give some ideas regarding what looks like the digital pedagogies. So and if you are interested and you can look at the article. So and the recent year and recently and that we have attention. This is the test of the fire and voice alarm system. There is no need to take any action. I think we will just wait a minute. Well I think there's a fire alarm gone off where Yaston's presenting. So if we just hold on a few minutes it'll come back again. Yeah there was no need to take any action apparently. Yes. So maybe Peter or Navi maybe you want to add something here in the gap. They're probably on the campus as well. Oh of course yes. That doesn't last too long. Oh yes you should know yes being Ulster that's fire. Sorry about the interruption. So this is our routine test. OK sorry. Attention please attention please. The test of the fire and voice alarm system has now been completed. Attention please attention please. The test of the fire and voice alarm system has now been completed. Would we have the slides after the presentation. Yes. OK. Thank you. So sorry about this. Giving the number of challenges and basically we lead to some kind of hybrid and the teaching and the learning mode at the moment but not the hybrid basically and come back to the physical campus teaching. So and however during this period of time so teaching a number of stuff and the student already got used this kind of either online and hybrid teaching environment. So and that leads to another question regarding what is the student engagement. So how we define the student engagement. So now let me to give you one first case study to talk about our experience regarding the how to analyze the student engagement. So in 19 in 2013 we organized one workshop mainly dedicated to the understanding relationship between student attendance with attendance. So in that workshop we invited one kind of the speaker on the leaves the burn from Durham University of Durham. She did one continuous study and she is the author of the paper on attendance, attainment and the final year study. So this published 2006. When we invited her she that's already six or five years past and she continually working on this topic and the ring gave us a keynote. So from his keynote he gave some bully point and from his her talk she still think there's a very strong correlation between attendance attainment and also student attendance gradually reduced. So for example from the semester one and until the semester two. So this gradually and gradually reduced and also so on the morning time per hapus is the worst time for student attendance and also if we upload the material on the beforehand per hapus that will affect student attendance as well. So from her study and she found so generally on student attendance for attendance increased improved and in past five years. So that's the her and the conclusion on from his keynote talk. So we did something similar study and then we reveal the literature and try to see what is the best way to understand the student engagement. So on the cool published one paper in 2007. So he divided the student engagement into different category two main categories and the first one called the behavior category and basically this kind of the attendance this kind of engagement we call the engagement can be and captured captured for example physical class attendance and how many times of students can access blackboard. So another category called the emotion category and but in that paper he did not give any solution any suggestion and how we are able to capture this kind engagement. So and based on our research and we tried to use some way to see okay so what is the possibility for us to incorporate our research into and this and the issue try to capture some kind of the student engagement. So we come up with a sentiment analysis. This is our resort area. So what is the sentiment analysis? Sentiment analysis sometimes also called subject analysis or opening mining. Basically this kind of technology can be used to analyze large amount of text document and identify the people openings or attitude to against a different subject or topics. So how we can use this one. So basically we can use this one try to help us try to help us to analyze some fun feedback document for example and the look at the whole student react to this kind of feedback document and then use some kind of student action to understand the emotion attitude of the student engagement. So this is the main approach and we used. So here's some bullet point under what we are interested basically we're interested to understand the learning experience and their attitude to particular program or module actually. The second part we try to understand because the feedback sometimes is very sensitive for students as well. So the reaction on the feedback is very important information for us. So how we are able to capture the student action from the feedback that is a very important part of Hoha's as well. So based on the some kind of analysis results and then we can see can we use some the auto come to group student into the different the group and then see some student perhaps the need to take a little bit of early invention in order to improve their the learning and learning auto come. So that's the basically the kind of the motivation and what how motivation for this piece of the work. So we have a fun system. This one was divided by our colleague Peter. Peter is here. So if you have any questions and so you can ask these some questions Peter. So fun system basically is able to convert the feedback into files and each file will be named into email student email address and then release all of the files into students with a unique and a hyperlink. So the system is able to record the two time stats. One when the feedback released to student and when student click the feedback. So this system has collected large number of module data and we just use the file years the feedback data and then started to do some pre-processing by end of the pre-processing and we have 3,300 feedback data for general study. In order to narrow down the scope of the study and we clearly to see 3 modules and 3 modules and across the different years and from year one to year four and try to understand the student and the reactions to the feedback. So this is one example and just mentioned. So here the system gives us the when the feedback file has been released and this time stamp clearly indicates when student clicked on the feedback. So we just based on the two time stamps try to understand so-called the emotion and the engagement of student. So what do we did? So we collect the data and then we project the data into two dimensional representation and then we divide this two dimensional into four areas. So first areas indicate the student's past but with less attendance. So under this X axis represent so how much class the student physically attended. So Y axis represents the mark the student achieved. So from here you can see under this part basically the past with less attendance and this one is the past with a good attendance. This part indicates the past and with less attendance and this part and indicate the field with less attendance. Okay so and the four different areas can be used a different purpose of the phone on our study for our study in order to improve our teaching and into the learning as well. So we do the further analysis use the sentiment analysis. We look at the how many days and the student collect their feedback and then we divide this the time into on the four categories and the first category and within the five days for example. The second within the file to 15 days and the 20 days will be the mutual and after 20 days will be the potential risk and the risk. So and we use this one to identify basically five group of students and then the five group of students give us some kind of ideas and how what kind of actions should we take. So this is the first part and our research results. So um so how we use this one to support our students. So for example once we have the clear the the group of student and the人 we can take different actions and in particular for example if we have two assignments and student reactions to further assessment not good within the potential risk area and categories and then we can take some actions take some actions okay. So and another part because this system is ongoing system so and widely used within the university and and Peter also have some ideas regarding and also based on the some feedbacks regarding what is the best time for us to release for example feedback to students. How this time will affect the student reaction to the feedback. I think that also is the important part for our next step research. So this is the first case study. If you have any questions and so and we can help discussion by and of my presentation and meanwhile so you can and put some questions into the chat box as well. So the second part I am going to talk about is the challenge of the chat GTP in assignment. So this one basically based on our MIC course I'm the module co-ordinator of the deep learning and it's quite a lot of research on the generative AI. The chat GTP is one of the topic of our research. So I think the chat GTP already not new because it's released at the end of the last year and already raised a lot of issues and the interest to the different communities. Basically we can use the chat GTP to generate different things. For example we can use the chat GTP to generate the code generate the text and the images and the 3D models and even include the music as well. So and the chat GTP basically based on the data the data come from the different sources and come mainly come from internet and come from libraries as well and they based on the data to train the model and use that model to do different generation of the image code the music as well. So because it can be used to generate different things naturally and we can use from teaching perspectives we can use and either for example help us to generate some assignment, generate the feedback to student the coursework and the student also can use the chat GTP to generate answers to their coursework for example. So here I'll give you a little bit of the background of one one module and a couple of deeper learning. This is a standard 20 credit module and teaching hours and basically is this 24 lectures and 30 practical sessions and six hours for tutorials. Independent study is 140 so because there's a standard model module so there's a 200 hours module. So we look at the assessment because this is mainly created during pandanum period this 100% of coursework. So the coursework divide the two part the first part is the writing assignment take 50% and the second part also is the writing assignment. So and I give you an example about the first assignment so the second assignment okay so this is the second assignment so and clearly to see what is the objective of this one and there's some specific task and the required student to complete. In order to understand how students use this one or use the GTP to complete this this kind of assignment and I invited two students from this AI course and one student and he wrote the beginning of this year Mohammed and the header ceiling. So and the second one and Martin Ousserman so he is from Egypt he involved the module study from last year last year okay two students and have experience about the coursework and also have experience of using chart GPT. So the both students confirm they spend less than two hours to complete this coursework and they also indicate if students don't have any experience about the use the chart GPT so under the purpose where you use a few hours to complete. So based on the result and the reason we see so under the result and perhaps the kind gets the 55 even 60% of the mark. So look at the original specification of this as a coursework so how much time we expect the student to spend on this one. So for example we expect students spend around the 40 hours but now you can see students only use the two hours to complete this one okay. So that's really the significant challenges for us so when we divide our coursework on the assignment for students so under what kind of style or the what kind of way should we use what kind of way should we use. So here's the answers from students and you can see students directly and put the requirement of the assignment into the chart GPT and use this as the copter prompt. And after that student can get the answer like this this is the answer from the chart GPT okay and the student also can require the code as well so and based on what I will get so and students can generate the prompt to generate the code as well. So a student did not tell me this code is working directly work or not but a student that tell me so they did not spend much time to make this one work okay. So what we learned so chart GPD under provide the good opportunities not only from the teaching perspective and also from the learning perspective as well so and the student can use the chart GPD to and use their own way to read the questions and then to get answers and then they can read Italy and then read the questions to chart GPD and then get the answers. It doesn't matter what time they want to use it as long as they have questions and they can use the chart GPD. So also for example for international students if they have language barriers they can use the chart GPD to have them to do different things in particular help them to complete and their assignment and the help to prepare the examination and the coursework that's what I already demonstrated so that's the opportunities but what is the the challenge is the complex cities and the chart GPD so and we can see the main objective of the coursework or assignment will be check a exam student the critical thinking and the problem solving skills. However if a student use the chart GPD and to create answers use the chart GPD to create some ideas so that means we are not able to achieve under what we pre-defined okay so and also if a student directly use the chart GPD as an answer so that involves the plagiarism also involves the misconduct as well related to the academic integrity. So student can repeatedly use this one to rely on use rely on chart GPD so that promotes the student's lazy. More things so chart GPD also creates some kind of disruptions to our teaching as well to our teaching as well so we initially design the our coursework with specifying the objectives however so when student use this kind of AI tools so and they can get different answers so and sometimes the general answer even challenging to us as well challenging to us as well so and something related to I think issues and the relate relate on the rely on tech technologies to do different things basically and that's the the challenges and from the generative AI tools when students use their use this kind tool to help them and do learning to learning. So we also noticed some new advice from the quality assurance agency in the UK and so under this is a basically and is kind of encouragement to to encourage and from teaching perspective from learning perspective to use this kind of new technologies so and for example the system statement clearly to see chart GPD is very useful and it provides a lot of opportunities for from the employee employability perspectives and also and pose a lot of challenges from the teaching perspective as well so they also create the one seven principles regarding the how to achieve the academic integrity integrity so and really regarding the this kind of things actually is the whole community rather than individuals rather than individuals so and I provide the link here and I provide the link here so and that you can have a look you can have a look so basically that is the some kinds of the guidelines from the QAA QAA so also from the different sources for example different units will say how different the based on QAA creates their own guidelines as well so um but so far on the way down to how clear the the solutions and the sort of way encourage the has to use the chat use encourage a student to use chance to complete their coursework or completed their teaching tasks or sort of way encourage and the complete stop on the student using this kind of tools so and now the come to the the summary of my presentation basically as a lecture so we are facing new set of challenges which I already mentioned at the beginning so and what extent so do we do bring the research into the teaching and this is the the big topic I think and today I don't have time to elaborate on our experience I hope and that we have a full opportunity so give you some detail so from the curatorial design perspectives and from the employability perspectives and why we think the incorporate the research into teaching practice is the more important so and we need to learn adopt to adopt a new emerging technologies for example particular from the generative AI chat gtp which I already demonstrated and also giving current the technology enhanced learning environment so and how we understand the student engagement that will be the important part for us as well so and there are a number of ways to try to capture student engage and with their program study but I think from our perspective we think is really challenging to understand the emotion the engagement for student the learning behavior learning behavior so but the important part where we are here is and we need to put a student in trun central and this is a guideline there's a strategy of the university university so and we have to understand the student's digital learning experience in order to put that in the central okay so this is an enjoyable my presentation so thank you very much for your pensions follow you hello hi thank you for the presentation very nice theme to start with because that's the way we seem to be going for the rest for the rest of these sessions so there's been a few questions come up in the chat here and there Elizabeth you were asking a few questions would you like to join in oh hello hello sorry I'm just trying to squirt the technology yeah um do you know you were talking about um that in your first case study the sentiment analysis which I actually had never come across but then I just googled it quickly loads of stuff came up so that just shows my ignorance but I really like that um focus on the emotional attitude of students so so that's very interesting um could you say a little bit more about that yeah um I think the sentiment analysis as I mentioned the sentiment analysis a sentiment analysis uh is a big restorative area in artificial intelligence so um basically underpin technology will be the motion learning so um so we can use motion learning to analyze a large amount of the text data to identify student attitude so um for our case study and actually we don't have a large amount of a text and what do we have we only have time stamps so we have um what time we upload feedback to students and what time as a system collected the student download their feedback okay so this is clearly this is a time stamp so and the some student for example we assume if some student engage their module study or program study and they should react the feedback very quickly so if a student not engage the module study or program study so a perhaps reaction to the feedback quite a slow quite a slow so that's the main important the the idea about the this approach so we try to understand their reaction and then we group them into five categories so for example for this part so basically student reaction good so and between under this period that period of time so we think no any potential risk for student studies however if student not collect their feedback and after 20 days so and then we if this is for first the piece of the course work and then we need to pay more attention on the second piece of work so we need to see why student reactions to slow and so this kind of slow give us the clear ideas and we need to put more attention to this cohort of students so I think that is the the important the outcome of our research 11 so it's okay thank you yeah that's kind of reading quite yeah this emotional analysis in that response which is an obvious one actually but yeah that's an interesting one to pick thank you so there's a few other little there's been quite a bit of chat going on in the background would anyone else like to add any questions at this stage yeah can I yeah so don't hear just an observation made on the previous slide you are showing here on the previous one where you showed the attendance rate and the marks yes looking at the scatterplot I would say there is no there's no correlation here between the attendance and the mark achieve would that be fair yeah I think this is our research result so um compare this research result with the least uh least burn from the term university so and it's quite different so and the way also this our continue study as well so um you can see actually this is a non-linear correlation so non-linear relationship so and from this one is a really difficult for us this clear correlation between students attendance and their achievement their achievement you're right so this gives us the idea about for example how we understand the student engagement can we see this group of students because you can see they did not attend the class very well but they still survived module study okay so what way they used what way they used to engage with module study I think this is a good question so that's why we come up with another part so um can we for example to clearly identify so-called emotion engagement under that group of students actually and if we use the emotion and the the engagement as the indicator indicator so and can we see that group of students actually emotionally engage module study very well so actually this our further research is the question question so because there are so many connections from different perspectives we haven't got a chance to look at how this is a group of students and uh the their emotion emotion engagement okay thank you um with someone else raise the hand I see oh yeah thank you um good afternoon thank you for sharing these interesting case studies and my question about the slide in the second case study about the prompt that you use in chat GPT yes do you provide the student with a certain standardized prompt or you give them the flexibility to use their own words so um because I try to understand the whole this is basically um I think in past half a year six months a lot of people to talk about the how to stop student use the chat GPT use whatever way uh stop student directly use the chat GPT to provide the answers to assignment so um the the case study mainly try to understand the student experience so I invited two students okay so um this student so uh you can see the the student two students actually study the deep learning model module module okay they have an experience on this course work on this course work so um but they also familiar with familiar with the chat GPT so I did not provide anything to student I just try to understand the student experience so what way student use uh to uh to generate answers to the assignment you can see from the I did not see I did not tell student how to do that and then you see student directly take the whatever I have in the course work as a prompt as a prompt I think that is a good indication so and if we clearly for example if when when we design course work so we need to take this factor into consideration try to award the student directly use under whatever we specified in course work as a prompt so I think that that will help is the one thing we need to learn we need to learn so um you can see student directly under use this one use what I specified in the in the course work under them to generate the answers and the two students use a similar way to generate the answers okay okay thank you okay anybody else got any burning questions at the moment and obviously all this will not be the last word we have on chat gtp I think it's just the stop way to go forward really um so I'm just going to remind people that we do have obviously some um the committee for alts special interest group is um open for invitations if you would like to join to become part of the committee there's a link in the chat and there's also it was a notice that went out by email this week so please feel free to nominate yourself I don't know whether Richard would like to say something at this point if he's still with me I'm still here um yeah I think it's it's just um Elizabeth I myself um I've come to the end of our three year term and we're looking for other um people to come along and and and take on the button and lead this um much needed interest group forward um I think regarded active run but specifically I think around the technology challenges we're facing with with artificial intelligence or generative AI et cetera um it's a fantastic time to come in and help shape the future direction of this group so really looking for anyone who's interested in becoming involved is chair co-chair secretary or even just member groups and FE um HE skill sectors this is a much frame much across multiple sectors group um so we're looking for anybody to come along and bring their experience and help share and shape where we're going to go forward with this but yeah so please do put your name into the hat you've got until January the fourth I think to do that the the I'll just post the link again on the chat for you so you can all see it yeah I okay okay yeah would you just do it Elizabeth yeah yeah it's all right more the more the merrier really um but yeah so um yeah please get involved if if you have the time or if anybody has ideas for webinars as well um you on our um forum there's there's other things you can submit and to get more involved with the group so thank you and thank you to Yassan and colleagues for today's presentation as well okay thank you thank you very much yeah I think that's all I think if no one's got any other questions then we can finish on time which is always good when people are in their lunch hour um and I say thank you very much for doing the presentation it's a very good way to start this new session for the for us and a reminder to if you wish to become part and also uh not forgetting to look out for the next notice for the next webinar so uh if if that's everything then we'll say goodbye