 Hello and welcome to the learning analytics tools course in NPTEL. This will be a two weeks course. I am Longmore Rajendra and as professor at education technology department of IIT Bombay. I offered the previous course called integration to learning analytics, four week course prior to this course. If you have done this course, you can see the same content or similar content for first two weeks. However, this course does not require any prerequest. So, even if you are not done the integration to learning analytics course, this course you can start. Due to web 2.0, the users can generate lot of data. For example, in social media networks like a Facebook, there are 1.56 billion users, monthly act users. And Instagram, QH, it is the lot of data has been generated. But what can we do with this data? So, how to use this data? For example, lot of organizations are trying to extract values or patterns from this data and using a lot of cloud computing, a lot of servers to try to understand what is a user behavior. For example, in Netflix, the video streaming websites like Netflix, the data for users viewing behavior, which movies they watch, the ratings, the likes and dislikes are used to recommend a next video for them. Similarly, in e-commerce websites like Amazon, the users based on users purchases, the system can recommend what will be the next product to purchase. In future, we can use this big data of DNA and the systems can come up with the cure for diseases like cancers or malaria, it is possible because the lot of data available, lot of DNA data of the humans and lot of medical data is available, we can use those data, we might cope with the cure for these diseases. Like in other domains, in education domain also, we have lot of data has been generated now. It is because of use of digital tools like learning management systems like a blackboard, Moodle or there is a lot of educational apps coming in, Google classrooms, massive online open courses like MOOCs and Coursera NPTEL or MIT Coursera, lot of these data has been generated. The users interaction with these interfaces has been collected and stored. What can we do with this data? Let us start with the first activity. Assume that you are a teacher, you may not be a teacher, but please assume that you are a teacher and teaching the same course for the same class or for last five years to the third year students. So, you are teaching say third year electronic devices course for last five years. So, you have data of the students for last five years, with their background, their profile, their performance in the exams, midterms, if you have done assignments, assignments course, you might have some feedback in the classroom, all this data you might have it. If you are to use this data to improve your teaching strategy or to improve the learners performance, what step will you take? Please pause this video and think about your answer, write down your answer, after writing down your answer and resume the video to continue. If your answers contain the words like, I will collect data, I will analyze data, I want to understand the learners learning or I want to improve my teaching performance, so that the learners can learn better. If you have these kind of words, then you are thinking in learning analytics. You already started thinking how to analyze the data to provide those analytics to your data and can improve the students learning performance. Let us see what is a formal definition of learning analytics. The learning analytics definition is still a debatable one, but from the existing resources, it can be the branch of analysis that makes use of students generated data for predicting educational outcome with the aim of tailoring education, with the aim of adapting the content so that students can learn better. Or from LAC, LAC is the organization which started organization called SOLAR, which starts a conference called LAC, the LAC 2011 conference and webpage they posted this as the learning analytics definition. Learning analytics is a measurement, collection, analysis and reporting of data about learners and their context for purpose of understanding and optimizing learning and the environments in which it occurs. What does it mean? It says learning analytics is measuring, collecting and analyzing the data, not just analyzing just for the analysis, also for reporting this data to a stakeholders of the LAC. Reporting about learners and their context, the sole purpose is to understand how learners learn in that environment and can we do something to improve the learning in that environment. In this definition, there are few terms. What is the purpose? The core purpose is to understand the learners learning process and help them so that they can learn better in the environment. Data collection, what data to collect, how to collect data, in which environment, what data to collect, how to use those data and analyze what to look for in that data and why we have to analyze. Also, if you want to report the data, to whom we have to report the data, so who are the stakeholders of LAC? For example, for the stakeholders can be educators that is the teacher or the instructor. The ability of a real-time insight into the performance of a learners, for example, if the teacher has a dashboard of the students working on a particular exam or particular learning environment, all the interaction is known given to the teachers by a dashboard. A teacher can have a real-time insight into the performance of learners, including students who are at risk and the teacher might know this particular students needs help, the teacher can go and help them or the teacher finds out old class having misconception a particular topic, the teacher can teach the topic in a better manner or give remedial content to that. So, it will be very useful for the educators like teachers or instructors. Similarly, for students, for example, if students receive information about their performance compared to the peers in the classroom or their progress compared to the peers in the classroom, that can help them to motivate and achieve their goal. For example, if a student answers a question wrongly, say the option be a student selected and he thinks the option might be correct, but if he showed to the student that in our class almost 40% of students selected this particular option and they are wrong, the student might feel I am not the only one who is given a wrong answer. So, everybody did not understand this concept. So, I can learn and the students get motivated to continue. If the student thinks I am the only one do not understand anything in the class, then they get demotivated, they might and they might not getting interested to continue further. Also, the student want to know how much progress in this particular course. If the student knows that I progress 40% of content in this course, I need to do another say 20-30%. The student can think of that saying that the student can think that okay, I need to put more effort to continue this course or the student can decide whether to continue the course or drop the course. So, the learning analytics can help the students motivate them and encourage them to complete the course and also help them to compare their performance with the peers. Finally, the learning analytics data can be helpful for the administrator to make decisions in the world today with a lot of competition and less number of budgets and the competition is coming in the administrator should take a decision that should provide optimal solution for them. Whether to run a course or not, if the administrator knows that particular course receives very less number of students and they know the why and if they can predict how many students will join the course in the next year, then they can make a decision whether to run a course or not they should convert the course into some other topic or something like that because it will save a lot of cost in the today's world because during the global competition in higher education, it is always useful to look at the data and make a wide stations. As you know that learning analytics is a relatively new field even compared to the other new fields like a machine learning or artificial intelligence. Since it is a very new field, there is no standard test book available for learning analytics. However, in this course, we will cover the basics of learning analytics in terms of analytics applied to education data. Then we will explain some tools useful for this course like tools to analyze the data. And we will also refer to some content from the test book. So, the test book for this course is and book of learning analytics from the solar community. The book is actually compilation of research articles explaining applications of LDA in a different fields. So, when you read this book, you can understand that every chapter introduces something about learning analytics and they will be applying LDA on some other topic. So, if you are interested in learning analytics, after this course, please take this book or read our papers in the recent conferences and you will understand the fields in learning analytics, then you can pick one field which you like to pursue your research interests in LDA. I will briefly describe the course outline in this video. In the week 1, we will be discussing what is learning analytics on academic analytics and what is the relation between this and with educational data mining. Very briefly, we will start over that. Then I will introduce what are the levels in learning analytics, the four levels in learning analytics with examples. In week 2, we will talk about the data collection, what data to collect in each environment. For data preprocessing, we might give you links to the external sources. So, we want you to go and study what is data preprocessing from the other resources and we might have a assimilation quiz based on that reading exercises. Also, in week 2, we will introduce a tool called WECA as a freely available open source tool everybody can use and will demonstrate the tool WECA which we will use in our course going further. Also, we will introduce ethics and data privacy when you collect data from the students, what are the ethics you should follow. In week 3, we will introduce a basics of machine learning. This course is not meant to teach machine learning. Also, the course may not involve mathematical details of each algorithm. So, this course is designed for anyone with a very little mathematics background to understand and how to collect data, how to start doing analysis. So, in week 3, we will introduce what is machine learning, what is supervised and unsupervised machine learning and very basics introduction to what are the metrics you should look for when you do the machine learning, apply the machine learning algorithms for your data. Also, in week 3, we will introduce a tool called Orange. That tool is for commercial purpose, it is not free. For academic users, if you have academic email ID, it is free to use. So, we will demonstrate that the tool called Orange. If you do not have access to that tool, you can continue using WECA in this course. In week 4, we will introduce descriptive analytics, how to describe the data and what is data visualization and how to look at the data, how the data is generated in a dashboard. And we can show using Excel or Google sheet, we can produce all these visualizations from the data we have. We do not need a sophisticated software for the research purpose. So, we can use Excel or Google sheet to visualize this data. And in a week 5, we will introduce another tool called ISAT. This tool is used for visualization also for the diagnosis purpose. This tool is developed by our department. So, we will demo the tool which will help you to visualize the data transfer from one stage to other stage in a different time periods. And we will talk about the diagnostic analytics and diagnostic analytics starts with correlation, regression, we will talk about correlations in week 5. In week 6, we will have a sequential pattern mining and the tool for sequential pattern mining will be described. The tool will be available free and we will upload the links to this tool and anyone can use this tool if the data is formatted in right format. We will show the demonstration of that. Also, we will introduce another tool called process mining, the tool called Pro-M. This is also freely available for educational purpose. So, anyone can use it if you have educational ID. Also, the Pro-M is available for everyone I think is open source. So, by week 6, we will be introducing several tools like a VECA, Orange, ISAT, SPM tool and Pro-M. So, five tools we plan to introduce in this course. So, after the mid semester break, in a week 7, we will talk about predictive analytics and we will talk very briefly about what are the features to select and doing a linear regression using the tool called VECA. So, VECA is again I mentioned it is open source, so anyone can use it. And in week 8, we will talk about decision tree and we will explain this with Orange. So, whenever we have a demonstration, when we are talking about explaining a particular algorithm with a tool, which means we will have a course assignments. So, you might be, you have to use the tool and we provide a data. Using that data, you have to predict something or you have to create something and report data's answers. So, yeah. So, when we are talking about this demonstration of tools, we will be having assignments in each week. And in week 8, we will have describe what is decision tree with Orange and we will talk about a naive base. As I mentioned this course, although it is not ML course, we will touch all these algorithms, which are very basics. We will explain the concepts, how it works, then we will also show how to use tools to execute these algorithms. In week 9, we will go for the unsupervised machine learning that is clustering and we will show what is clustering and with a demo. In week 10, we will jump to a different fields called test analytics or natural language processing applied for educational domain. So, we will show you how we can use test analytics to develop algorithm that can automatically grade students essays, a simple algorithm and also we will talk about the latest development in NLP that is called word embedding or word vectors. And in week 11, we will talk about multi-modal learning analytics that is how to collect data from multiple sensors like eye gaze data from eye trackers or facial expressions for using web cam, log data or some biosensor data like EEG or EMG or GSR data, how can you collect these data and how can you use these data, analyze these data to create a model. So, we will show a bit about what how to collect data, what is this data is used for and what is this model looks like. In the final week, we will teach about advanced topics in LA and what are the topics. It is basically topics covered in the handbook of LA also the latest topics published in LAC conference and EDM conferences. So, that is the course introduction. So, in this course, we briefly describe what is definition of LA or LA means you have to collect data, analyze data, then you have to report data to stakeholders to improve the performance of the students. Then we describe the course outline. The course outline involves five tools, a lot of assignments and a lot of data to collect and a lot of exercises in it. And that is all for the motivation video. This is the first video. Thank you for watching.