Tag Archives: LRMI

Big Data Analytics for K-12 Personalized Learning

KimJonesimageBy Kim Jones, CEO, Curriki

There is a rush, perhaps a gold rush, underway in efforts to leverage Big Data analytics to improve K-12 educational results.

A story from Reuters http://www.reuters.com/article/2013/03/03/us-education-database-idUSBRE92204W20130303
reports that a new $100 million data warehouse has been built in the U.S. to monitor academic achievements of public school students, from kindergarten through grade 12. The article states that “The database is a joint project of the Bill & Melinda Gates Foundation, which provided most of the funding, the Carnegie Corporation of New York and school officials from several states.”

Data_warehouse_overview

Already 9 states in the U.S. are participating to some degree, and two of these, New York and Louisiana, are planning to provide all or most of their student records into the data warehouse. The database includes students names and addresses, and other personal information.

A non-profit named inBloom, Inc. (previously called the Shared Learning Collective) has been established to operate the database, which already contains millions of student files. It is a cloud-based data warehouse for student data such as grades, test results, assessments, standards met, behavior and attendance. It is not a repository for digital content, but will contain links to content, leveraging the Learning Resource Metadata Initiative methodology. The stated goal for the data warehouse is to support more personalized learning.

inBloom.logo

Here you can watch a video that introduces inBloom’s vision for customized teaching and learning:
https://www.youtube.com/watch?feature=player_embedded&v=gHjbdpXohk0#!

A very large amount of data is gathered by school districts and states, but resides in many separate databases and is not cross-correlated or well-analyzed. This project aims to change that. See this Mindshift article:
http://blogs.kqed.org/mindshift/2012/07/how-will-student-data-be-used/
Pulling data together in this way is intended not only to make learning more customized, but also to make it easier to track results as students move from district to district or even state to state. In addition, it will allow data analysis on student achievement to be linked to various learning resources, potentially including the 46,000+ resources on Curriki.

Educational software suppliers are excited by the opportunity to mine the database and better determine what educational products to develop, including educational games and other digital learning products, lesson plans, and reports.

The goals of the project are laudable, but not surprisingly, given the sensitive nature of the data, privacy and security concerns are being raised. inBloom states on their web site: “We recognize the sensitivity of storing student data and place the utmost importance on the privacy and security of that data.” Parents in some states are already raising concerns about potential data leakage. While the data warehouse contains essentially the same data already held in school district databases, it now becomes available to a wider audience, including educators in other states than the original source for the data. And educational content vendors are requesting access to test data, for example. Presumably this will be made anonymous when supplied. Ownership of the data is retained by the states and districts that supply it. But some organizations, including the PTA and ACLU, are already asking “What are the remedies if and when data leaks?”.

We’d be interested to hear what you think about this project. What benefits and risks to you see in amassing a well-integrated and analyzable database of K-12 student achievement? How would you want to see this data being used to improve learning outcomes? Do the benefits outweigh the risks? Please provide comments.

Video

LRMI will provide better Learning Resource discovery

JoshuaMarks3By Joshua Marks, CTO, Curriki

LRMI? Why should you care? While LRMI, when pronounced like a name (Ler-me), might sound like the lead singer for Motörhead, it is really an acronym for the “Learning Resource Metadata Initiative”. Now you might be more interested in rock trios than educational tech acronyms, but if you are an educator or student, you will find LRMI much more enlightening and helpful — and perhaps you will spend less time banging your head against a wall. This is why you should care.

The World Wide Web is filled with billions of documents, images, media, articles and content of all sorts. Search engines like Google and Bing do a great job of helping you find things you are likely interested in using keywords and phrases. This is particularly true for things you might buy, or people or places you want to learn about. However, anyone who has looked for learning resources on the Web sees that only a small fraction of the documents Google suggests are specifically useful in teaching and learning a topic at a given level. If you narrow your search with terms like “lesson plan” or “textbook” you get more relevant results, but still miss a lot. And you still might find lessons or books that are at the completely wrong level or skill. Recently released research underscores this discovery challenge and highlights the need for a better way to identify and find web based learning resources: http://www.lrmi.net/survey-results-show-need-for-more-targeted-results-when-searching-online-for-learning-resources.

When I was asked to join the Technical Working Group for LRMI, I believed that if LRMI simply gave us a way to search only those sites that contained content intended as learning resources, we would be in a better world. But the real task the group took on was “How can we provide a simple way to identify (tag) all Web pages that contain content intended for instructional use as  ‘Learning resources’ and specifically included information about the subject, topic, how it is to be used and how it might align to local learning standards or the Common Core currently being adopted in the United States?” What if you could search for just those pages that contain activities appropriate for 10-year-olds, videos that illustrate photosynthesis, or some other specific topic?

The technical working group has completed its work and the V1.0 specification (http://www.lrmi.net/the-specification) is being adopted both by Schema.org, which is a project led by all the big search engines (Google, Yahoo, Bing, etc.), and the W3C (the group responsible for all Web standards). The next step is to get all the Web masters and publishers in the world to use it. This is where you come in. The more you tell the Web sites you use that you want them to support and tag the content with LRMI tags, the better  job Google, Curriki, Bing and everyone else will do of locating the content you are interested in, without all the frustration and bad matches.

You can see this kind of tagging in action in other domains right now. For example, if you search Google for “chicken salad recipe” then under “More” select “recipes”, and then select “Search tools” you will see a bunch of very specific recipe filters like ingredients, calories, and cooking time. (It is way too hard to find in Google right now, but really helps finding a recipe.) You can image a similar set of filters for Learning Resources. In fact, you can see this functionality right now when you search for content in the Curriki community in Advanced Search, where you can filter results by subject, topic, level, language, instructional component type, rating, reviews and all other Curriki-specific metadata (metadata being information that describes content.) This is what LRMI will standardize for the entire web.

This is just part of the story.LRMI has been used by another project called The Learning Registry (http://www.learningregistry.org/), which in turn is part of a U.S. initiative called the Shared Learning Infrastructure (SLI) (http://blogs.kqed.org/mindshift/tag/shared-learning-infrastructure/). These extensions of LRMI are supported by both the $4.35 Billion Federal Race-to-the-Top grant program and the Gates Foundation. When taken together, these independent but interlinked initiatives are finally creating a critical mass of movement toward common digital material formats and an information infrastructure that enables a full transition to personalized digital instructional delivery. In fact, without these standards and the central role LRMI plays in providing a common way to define the use and target for learning assets, the dream of personalized dynamic learning systems would not be possible.

So stop banging your head looking for the instructional needle in the information haystack known as the World Wide Web. Tell everyone who publishes learning resources we need them to use and support LRMI!