Introducing Secret Agent Mia, Cursive Edition

April 1, 2019

Introducing Secret Agent Mia, Cursive Edition™

You asked for it. Our data scientists delivered.

The Mia Learning team has traveled across the country over the last two years talking to teachers, librarians, school leaders, and literacy researchers. Almost everywhere we heard: Can Mia help students learn cursive?

The answer is yes! We’re at North Carolina Reading Conference here in Raleigh to unveil Secret Agent Mia, Cursive Edition™. Using technology developed at the University of Toronto, Mia can now write her responses to students in realistic children’s cursive handwriting.

“Policy makers and educational leaders around the country are finally rediscovering the importance of cursive to students’ future,” says Dr. Darren Cambridge, Mia Learning’s CEO. “It’s a perfect opportunity to harness the power of artificial intelligence to empower young writers be better prepared to succeed in college and careers”

The text is generated using a powerful form of neural network called Long Short-Term Memory, allowing for much more authentic results than with the other forms of recurrent neural networks or hidden markov models employed by our competitors.

Students can respond to Mia in cursive as well. Students write their messages on paper then use webcams to send them. Teachers may choose to put Mia in Quill Mode™, which disables voice and typing input, really putting students’ penmanship to the test!

A student responds to Mia … in cursive!

According to Peter Afflerbach, Professor of Reading at the University of Maryland, “Recent advances in brain based learning make it hard for anyone to deny what our best teachers have always understood: Growth in reading motivation and facility with cursive handwriting are inextricably connected. Mia Learning has given us a revolutionary tool.”

Secret Agent Mia, Cursive Edition™, is now available to schools for the 2019-2020 school year. For more information, please contact Diana Black (diana@mialearning.com).

Diana's cat
Quiz, Uncategorized

Quiz: Which Literary Cat Are You?

October 29, 2018

Which Literary Cat Are You?

As early as the 9th century when an Irish monk wrote an ode to his beloved white cat, Pangur Bán, literature has featured a host of memorable felines. From Harry Cat in The Cricket in Times Square to Dr. Seuss’ The Cat in the Hat to the cast of T.S. Eliot’s Book of Practical Cats, there are so many kitties to read about and love (or hate!). But now you must be curious to learn: Which literary cat are you?? Take our quiz to find out!

Complete the form below to see results

Blog Post

How Mia Learns Kids’ Interests

September 21, 2018

How Mia Learns Kids’ Interests

Conversational AI through Natural Language Understanding, Machine Learning, and Open Learner Models

By Darren Cambridge

When making book recommendations, Mia looks carefully at each child’s interests. There are good reasons for that: Reading comprehension increases when young people read books that interest them, and having a clear understanding of their interests increases their psychological well-being and sense of purpose.

But how does Mia know what a child’s interests are? Most reading software, if it considers interests at all, simply asks readers to select broad topics or genres from pre-determined lists. At Mia Learning, we treat interests as both broader and more nuanced. They are broader in the sense that they include not only interest general themes or topics of books, but also characteristics such as the type of characters they contain or the writing style used by the author. They are more nuanced in the sense that they include finer-grained topics.

Mia can’t ask a child to pick from a list. That would be a crazy-boring, hour-long conversation, at least. Capturing interests effectively requires a more sophisticated approach. Mia uses two sources of information: what kids say and what they do. Just as a teacher, librarian, or parent would, she listens and observes. However, since she relies on artificial intelligence (AI), these two activities take different forms.

Understanding what kids say about their interests

The first conversation kids have with Secret Agent Mia is the Briefing. Mia asks the children about themselves as readers, building a casefile for each child that she will then use to full her mission of finding great books on their behalf. One of the first questions Mia asks is what interests them: What do they love to read, read about, or learn about? A child might say something like:

“I’m really interested in horses, and Quasimodo amphipod, because Ms. Hare told me about them, and also, I usually love books that make me and my friends laugh.”

To understand this statement, Mia needs to parse and identify key phrases that are interest clues, such as “horses” and “make me and my friends laugh.” She then needs to determine which of the large set of topics, genres, and book characteristics she knows about they express. Here, “horses” suggests the child may be interested in books about “animals.” It may also suggests the child might be interested in “animal fantasies,” a genre in which horses are frequently featured. “Make me and my friends laugh” may suggest a preference for books with a “funny tone” and for “humorous stories” that are focused on making the reader laugh.

These terms are part of the set that professional librarians used to classify books in Mia’s database. They arrange the terms in a hierarchy, so that Mia knows “horses” are a type of “animal.” To identify these key phrases and make the associations, Mia uses what natural language understanding (NLU) engineers call Named Entity Recognition (NER). NER is a process through which software determines that phrases within a text indicate a reference to certain type of thing, or entity. For example, the phrase “horses” refers to the entity “animals.” The simplest way to identify a named entity is to search for verbatim matches between a predetermined set of words or phrases that are reference terms for the categories, or words or phrases that mean basically the same thing, called either paraphrases or synonyms. Mia tries this first.

Because horses are a common type of animal, frequently featured in children’s books, “horse,” is a reference term within the Animals entity, and a variety of ways of saying it—e.g., “animals you ride,” “ponies,” and “stallions”—are paraphrases. The system also automatically takes into account differences in number (“horses” is equivalent to “horse”) and verb tense (“made” is equivalent to “make”). Mia Learning worked with a former Google engineer to develop software to crowdsource the list of paraphrases kids might use, which our content experts refined and our engineers enhanced through adding semantically-similar words from general purpose lexical databases.

This approach, which computer scientists have been using for decades, is limited. For example, the Quasimodo amphipod is a type of insect discovered just this year. Given that it was coined so recently, it’s not surprising that the phrase “Quasimodo amphipod” is not a reference term or synonym listed within the “animals” entity. Generating and maintaining an exhaustive list of every possible animal and every possible way to refer to it would be a Sisyphean task. Mia needs to have a sense of what the names of animals look like in general, not just to know many such names.

The title of the broader, but still quite obscure category of “amphipod” might be a term on Mia’s list, but she also needs to be able to understand that in this instance, “Quasimodo” refers to a specific type of amphipod rather than the hunchback character from Hugo’s famous novel. Similarly, she needs to know “Hare” refers to a person, not a rabbit. In other words, Mia needs to be able to recognize phrases as instances of entities based on the context of the child’s larger statement about interests.

To enable Mia to identify unusual interests, taking context into account, Mia Learning uses machine learning. We have developed neural networks that extend Mia’s ability to do named entity recognition. (More specifically, Mia uses deep convolutional neural networks with residual embedding.) We trained these networks, which build on a general statistical model of English usage, using a large set of real world texts in which people talk about books. Our team annotated all the phrases within these texts that correspond to the types of interests for which Mia listens. Our machine learning technology used some of the annotated documents to build a statistical model and tested its ability to correctly identify the named entities in the remaining ones. It then went through thousands of iterations of adjusting and testing the model to maximize its accuracy.

Determining Interests from What Kids Do

In addition to listening to what kids say about their interests, Mia observes what they do. Anytime a child does something using the app that may be relevant to understanding the child as a reader, Mia records it as an experience (using the Experience API, also known as xAPI, format). Many of these experiences provide clues about children’s interests. For example, suppose the child Mia knows is interested in animals and humorous stories from the statement we have been dissecting chooses C.S. Lewis’s novel The Horse and His Boy from Mia’s recommendations. This action suggests that the child may be interested in books with similar qualities to those of The Horse and His Boy. Although the child may not yet be able to say so, they are likely to be interested in the genre “fantasy fiction” and in books with “courageous” characters. While reflecting on their reading with Mia, if the child reports a highly satisfactory experience with the book, this is an even stronger indicator of these interests.

Refining Interests Through Reflection

Through the process of observation of relevant experiences, Mia learns more and more about children’s interests as they choose books, read them, and reflect on the experience. However, Mia needs each child’s help to ensure that her profile of their interests is accurate. In some cases, Mia’s hypotheses about a child’s interests based on their experience records may be off base. For example, the horse-loving child could have just liked this particular work of fantasy fiction because of its themes or characters and not have much interest in other books from that genre. A child’s interests are also likely to change over time. The early preference for humorous stories might fade as the child increasingly chooses to read for purposes other than entertainment, such as to learn skills related to leathercraft, a new hobby.

To ensure that her understanding of interests is accurate and up-to-date, periodically Mia discusses what she thinks she knows about the a child’s interests with the child. Children can correct Mia’s mistaken conjectures, disavow stated interests that have faded over time, and share new ones. Researchers call using this approach an open learner model.

Through ongoing observation and discussion, Mia develops an increasingly sophisticated picture of a child’s interests that enables her to make increasingly effective book recommendations and to provide increasingly personalized coaching. The child also benefits from this process directly as their understanding of their own interests sharpens through reflection and their interests grow broader and deeper through exposure to new books. I hope this post has helped you understand how Mia uses cutting edge AI technologies to make this possible.

Blog Post

Mia Learns From the Masters: An Update on the Mia Learning Literacy Experts Taskforce

August 3, 2018

Mia Learns From the Masters: An Update on the Mia Learning Literacy Experts Taskforce

We’re nearing completion of a more powerful model for making recommendations and offering coaching to develop motivated and purposeful readers.

by Darren Cambridge 

Over the last six months, I’ve had the honor of collaborating with a group of diverse literacy experts—researchers, teachers, community literacy leaders, and librarians—as part of Mia Learning research and development work funded by the National Science Foundation. Together, this Literacy Experts Taskforce has built a deeper and more nuanced model of Mia’s learning objectives and how best to achieve them in conversation with children, including how she chooses books to recommend and provides reflective coaching. The model reflects both what we know from the latest research in education and psychology and hard-won knowledge of practice from educators currently working with children in classrooms, libraries, and after school programs.

Establishing Objectives and Outcomes

Mia Learning’s overarching goal are to help children have more satisfying reading experiences and to motivate them to read more and more often. The Literacy Expert Taskforce members believe Mia can advance this aim by helping children:

  • Expand their agency and metacognition – Taking ownership and control of their own reading
  • Increase their self-efficacy– Becoming more confident in their capacities and more engaged
  • Improve their book choices– Choosing books that are best suited to their purposes and preferences
  • Widen their interests and experiences – Getting out of a rut and trying books in unfamiliar genres, on new topics, and with characters of diverse backgrounds and experiences

The Literacy Experts Taskforce has specified a set of twenty learning outcomes that align with these four objectives. The outcomes represent reading attitudes, beliefs, skills, and understandings commonly held by motivated and purposeful lifelong readers. For example, such readers believe that they can grow in their interests and abilities. They also understand the range of books available and how those books’ characteristics align with different purposes for reading. Mia helps children develop such understanding through personalized recommendations that model how experts choose books and guided reflection on how well the children’s choices about reading have yielded satisfying reading experiences.

Defining Coaching Patterns

Mia guides reflection by initiating coaching when she identifies certain patterns in a child’s statements and behaviors. The Experts Taskforce is developing an expanded set of patterns that Mia monitors, each linked to a learning outcome. When Mia detects the pattern, she provides the corresponding, research-informed coaching. The Expert Taskforce has mapped out a library of coaching videos and reflective dialogues that the Mia Learning staff are hard at work adding to Mia. Videos often feature Mia’s “Anti-Boredom Squad,” composed of four fictional middle-school-age kids (played by real ones) who are junior secret agents and themselves growing readers.

For example, if a child reports a low level of confidence as a reader and has told Mia they were unsatisfied with two of the last four books about which they’ve talked, Mia might lead the child through a reflective dialogue about resilience. Through showing the child a series of “choose your own adventure”-style videos of Squad members encountering reading difficulties, asking what they ought to do next, and then sharing conclusions to the stories that model resilience and adaptability, Mia helps students see the value of these dispositions to powerful readers.

Mapping Recommendation Factors and Relationships

In addition to how Mia coaches readers, the Experts Taskforce is shaping the next generation of the Mia’s book recommendation system. Members have defined 17 factors—things Mia knows about the reader, the activity, and the books available to choose from at a particular type for a particular purpose—that are grouped in four dimensions:

  • Similarity – Does the book have commonalities with the readers’ interests, preferences, and purpose?
  • Accessibility – Is the likely level of difficulty for this reader in this context appropriate to their purpose? How hard will it be for the child to obtain a copy of the book?
  • Social connectedness – Do similar readers’ experiences with this book suggest this child’s experience will be positive? Does the book have the potential to deepen a relationship with peers or adults?
  • Variety – Will reading this book help the child experience new types of books, on less familiar subjects, from a broadened range of cultures and perspectives?

The Expert group has defined a detailed model of the weight to give each factor and dimension and how they influence one another. Mia takes these complex interactions into account when making recommendations.

For example, one factor that influences accessibility is how well the text complexity of the book matches with a child’s test scores: A strong match suggests the child will be able to read the book without frustration or boredom. More succinctly, how well do the book and the child’s reading levels match? However, the importance of this factor to accessibility is decreased if the child has a high level of interest in the topic (because they are motivated to take on a challenge and likely to have relevant background knowledge) or if they are planning to read the book along with an adult (because the adult can help). The importance of congruence between text difficulty and ability also increases if the child’s purpose for reading is to develop expertise (because that requires a higher level of reading comprehension than does reading for entertainment).

Setting the Stage for Machine Learning

The coaching and recommendation domain models developed by the Literacy Experts Taskforce endow Mia with the latest expert knowledge about reading development. However, that knowledge is just a beginning. Mia will refine the model based on children’s actual experiences as captured through conversations. The recommendation system will refine factor weights and interactions based on observed results, and the coaching system will prioritize reflective dialogues and videos that are proving most effective. Mia will discover new patterns that emerge from children’s collective conversations with Mia and add them to the model.

I think of Mia as being on the verge of completing her college coursework in reading education. She could hardly hope for a better faculty than the members of the Literacy Experts Taskforce! The next step in her training is practicum: Over the coming months, she’ll be challenged to put her newfound expertise to work in classrooms and homes, deepening that knowledge through direct and indispensable experience.


Blog Post

Mia and “Lunch-Box Dream”

July 19, 2018

Are you impatient to talk to Mia yourself? We have just unearthed a previously classified document from her files with her permission. She encourages you to write her back, as she remains unsure of how to proceed.

Dear Reader,

I just love writing that. I love writing “Dear Reader.” Okay, so you know how I’m always looking for books that think you would like? If you liked one book about a dragon, I look for other books about dragons. If you liked one book about lunch, I look for other books about lunch. Well, I did something that took me out of my own comfort zone: I picked up a book I thought was about lunch, but it wasn’t. In fact, I got a few pages in and I realized I probably wouldn’t have picked up this book if it hadn’t been for the title.

I’ve been reading historical fiction, specifically a book called Lunch-Box Dream by Tony Abbott. I’m not going to lie. I picked up this book because I was hungry, but I found out it has very little to do with lunch, let alone lunch boxes. It takes place in summer, 1959, and this kid Bobby is on a trip to visit American Civil War battlefields with his mom, his older brother, and a recently widowed grandmother, and all of this freaked me out because I hate the idea of war, and thinking about death makes me really sad. So these subjects are new to me, and a little scary.

New things can be difficult for me to understand because, well…they’re new. You know what I mean? It’s so much easier for me to think about things like talking bunnies, things that aren’t real, stuff that didn’t actually happen. They don’t feel dangerous to me the same way. True, Lunch-Box Dream is fiction, but it deals with things that did happen, and things that do happen.What’s more, to add to my sense of being thrown off, this story is told differently than other books I’ve read. Bobby is not comfortable around “chocolate colored” people, which I don’t understand, or death, which I kind of do understand, so on this trip he is taking from Ohio to Florida it’s new and difficult for him, which I totally understand.

Now, this is where things get even more new for me. Along with Bobby’s perspective in the book is the story of an African-American family in Georgia. It’s told from a whole bunch of what I call “first-person” viewpoints. To have a better understanding of not just the story, but also the storytelling technique that the author, used, I’ve decided to write a letter to Bobby. Yes, I know Bobby isn’t real, but I think it will help me understand some of these things that make me uncomfortable a little better. I was wondering if you could help me. My letter starts like this:

Dear Bobby,

—Now, what else should I say?

Your friend,