I Read Arabic

Problem we are trying to tackle:

Arabic is the fifth most spoken language globally and the second hardest to learn. According to the World Bank 50% of children in MENA suffer from learning poverty, meaning they cannot read or comprehend an age-appropriate text (World Bank, 2021). This is further evidenced by very low scores on global standardized literacy tests of participating Arab countries (PISA & PIRLS). 

Why?

Reason vary: for one Arabic is hard to learn in that it is a diglossic language where the spoken language is different from the written language with only 60-70% overlap (World Bank, 2021).  Furthermore, public schools (which is where the majority of students go) are overcrowded and underfunded with poorly qualified teachers and dated curricula. Regional conflicts in the region for the past 10 years and more has also led to the displacement of millions of children who have lost years of schooling.  Finally, there is a huge lack of engaging interactive Arabic learning platforms that actually measure and improve outcomes. The result: same, a generation with a severe learning disadvantage.

Why literacy?

When children read by grade four, they become lifelong learners, and when they don’t, they suffer in all other subjects as well. Poor literacy leads to high drop outs and eventually high unemployment (a major cause of extremism), early marriage and poverty. Given that the ability to read and understand a simple text provides a foundation for further education, it is imperative that children get the language and cognition development they need in the early grades of schooling (Duke, 2019). Progress toward eliminating this learning crisis necessitates leveraging technology to design learning experiences that are both equitable and impactful on learning outcomes. This is where we come in, and this is what we do: creating Arabic language learning platforms to improve Arabic literacy.

We create online gamified reward platforms designed to improved learning outcomes for over 30 million children in 11 countries in the MENA region covering students in private, public and remedial schools that serve refugee and displaced students.

Little Thinking Minds is committed to improving Arabic literacy and language learning outcomes for millions of children in the Middle East and North Africa by providing a levelled learning journey with age appropriate, engaging, interactive content and learning tools. We create subscription based educational resources and platforms (app and web based) aimed at improving learning outcomes and enhancing children's literacy and fluency skills. The apps include very rich content for the teacher and student and includes Arabic e-books, audiobooks, videos, games and interactive activities that are rich and aligned with regional curricula and learning outcomes. The applications offer students a gamified reward-based experience. For teachers and schools, our applications offer a fully comprehensive Arabic language learning experience suited for both inside and out of the classroom with endless resources to support the learning process and deliver real results. Additionally, our platforms are complemented by a teacher dashboard that enables teachers to track their students’ progress (level, fluency, time on task) and identify weaknesses and help them improve their results through remediation

Our future plan is to develop and upgrade our existing product portfolio integrating AI to allow for a more effective language learning experience. The new tool will leverage Machine Learning (ML) algorithms to personalize and adapt the learning experiences in real-time to fit learners’ strengths, needs, goals, and pace of learning (i.e., personalized and adaptive learning, PAL). This PAL tool will capture data to quantify the main drivers associated with Arabic literacy development. The PAL tool will aggregate and analyze this data to make continual adjustments to learners' learning pathways. Specifically, this tool will identify learning content for skills development at the individual level and build a model that leverages multimodal learning. This tool will also enable a dynamic multimedia approach to support learner motivation and persistence along curricular modules. The primary output will be a field-tested product that can be used in classroom and at home settings. This design solution will serve as a resource for multiple stakeholders to improve learners’ Arabic literacy outcomes through automated instructional support and a real-time feedback mechanism to better understand learner learning processes.

Our solution serves school aged children between 4- 18 who are either in formal schooling or not. The app provides a personalized journey where users embark on a journey where they read/ hear books, record their voice and answer questions aligned with learning outcomes. They can also watch short animated educational videos aligned with regional curricula. As children progress through the levels they go through a series of tests (end of level tests and benchmark tests) to assess how they are progressing. Teachers can assign specific tasks for struggling readers through a remediation feature.

Our solution also serves Arabic language teachers who are often marginalized and struggle to find engaging resources for their students. Teachers have access to a rich library of hundreds of books, worksheets, videos, lessons plans and assessments which they can use in their classroom or assign as homework. They also have access to detailed reports measuring achievement of learning outcomes and identifying struggling readers.

The platform can also be accessed by Heads of Departments, Heads of Districts and Heads of School to monitor students and teacher's engagement on the platform. 

The co-founders of Little Thinking Minds bonded over a common frustration with the lack of engaging Arabic literacy resources for their own children. What started out as a fun project in Jordan mushroomed into a platform that reaches hundreds of thousands of children today.

We are a team of 60 passionate minds based in the Middle East; our headquarters are in Amman- Jordan (serving Jordan, Palestine, Iraq, Lebanon) and we have offices in Cairo, Egypt (serving North Africa), Riyadh, Saudi Arabia (serving Saudi Arabic) and Dubai, United Arab Emirates (serving the GCC).

Our team is divided into five departments:  tech, product, content, sales and marketing. Our team members include educators, writers, illustrators, developers, marketeers and passionate sales and customer support people.

Our team is 65% female and 50% mothers and 100% mission driven, all ardent believers in the company’s mission to improve Arabic literacy.

Being in four countries we work very closely with our target segment conducting multiple visits a year as well as follow ups on zoom calls where available. We also conduct continuous focus groups with our user base of teachers, students and parents and send monthly surveys. Additionally, we invest in data collection and data analytics where we study user behavior and iterate on our product accordingly.

Our company has been recognized regionally and globally, the most recent being selected in 2020 and 2021 as part of the 2021 Elite 200 (GSV Cup Competition) representing the top “Pre-K to Gray” EdTech startups across the globe with a mission to ensure that children have access to a better future through education.

Monitoring & Evaluation

Our current platform works on accelerating K-12 Arabic literacy by providing an engaging and individualized learning experience. We are working on developing a Machine Learning tool to also for a personalized and adaptive journey. Our platform is evidence based; a third-party impact evaluation has shown statistically significant improvement in literacy skills specifically syllable identification, reading comprehension and oral reading fluency (ORF). This is based on a third-party impact evaluation conducted by School to School Research (a San Francisco based research company), part of the All Children Grand Challenge for Development (funded by USAID, Australian Aid and World Vision). We were one of a small number of global winners to be awarded this Grand Challenge focused on innovation in literacy.  We applied the solution to 10 public schools comparing treatment to control groups. To read more please click here.

(Full study: https://allchildrenreading.org/wordpress/wp-content/uploads/2017/10/Little-Thinking-Minds-End-of-Project-Report.pdf)

We know our individualized solution works. What we want to do now it to make it adaptive and personalized and machine learning to give users/ teachers/ parents more insights on their learning journey.

We believe this will be catalytic and will change how users learn how to read and transform them into users who want to read to learn.

Plan: Make journey personalized and adaptive

All too often, curriculum in the region promote and reinforce grammatical and spelling rules in a rote and abstract way that discourages children from making up words or misspelling, while traditional whole-word approaches to teaching the alphabetic system prevail (Saiegh-Haddad & Schiff, 2016). Research indicates that a scaffolded and differentiated learning approach is more effective than these traditional approaches to literacy instruction (Bondie et al., 2019; Gregory & Chapman, 2012; Smale-Jacobse et al., 2019). However, tailoring instruction to each student’s needs and preferences can be a challenge to practice as it entails far more than the adaptation of curricula (Dosch & Zidon, 2014; Turner et al., 2017).

The PAL tool we will develop will be leveraged to tailor learning experiences to fit students’ strengths, needs, goals, and pace of learning while supporting and extending the capacity of teachers. This tool will span the entire range of early reading; it will begin with letter naming and sound recognition, to being able to read sentences and then books through connected text. The learning content will be presented in a fun and engaging way and will follow a gamified approach to assessment. This tool will aid teachers with the teaching-learning process by facilitating real-time monitoring of student/cohort progress while easing administrative tasks like grading reading exercises. It will also translate the performance data into actionable insights and nudges teachers to intervene when a student is struggling.

Plan: Create proper assessments and remediation

Formal instruction in the region is too reliant on a one-size-fits-all curriculum and teachers too often lack the ability to identify and respond to students’ needs (UNESCO, 2016). If learners' progress is not adequately assessed, they may end up moving through the curriculum without mastering the needed reading skills. Ultimately, this can impact their ability to learn other subjects, and negatively affect their life chances. Also, parents will not know when their child needs additional support, and ministries of education will not be able to assess the effectiveness of their education system.

To provide the best literacy instruction, learners need to progress in a logical and well-planned way while ensuring that their learning gaps are addressed. This, however, is a difficult undertaking in large classrooms. By leveraging technology and through developing an ML tool, this tool will place learners in the right reading level, allow them to learn in their preferred learning style, while identifying areas which the learner needs to master. Teachers will have a dedicated dashboard to assign tasks and view individual learning and class-based detailed reports highlighting the mastered skills and the challenging areas. Since parents are important facilitators of children’s learning, this tool will enable parents to take an active role in their child’s learning journey by providing them with detailed progress reports so they can monitor their performance and provide timely support when needed. 

Our goal is to create a solution that teaches Arabic reading. We want to make learning Arabic easier for students. We aim to create a solution that improves users language skills, improves their listening, reading, comprehension skills, increase their confidence and turns them into avid readers. Our goal is to help students become confident and avid readers, proficient in their mother tongue. Not only are we focused on how to improve reading technically but also through our 800 books which cover various topics from STEAM, mindfulness, gender equality, social emotional themes and SDGs ,we hope to expose the readers to global themes of inclusiveness and plant a seed for curiosity and love of learning.

We are currently measuring:

• Time spent on platform reading;

 (on average Arab children read 6 minutes a year according to the Arab Thought Foundation https://en.unesco.org/creativity/policy-monitoring-platform/arab-reading-challenge#:~:text=Reports%20and%20studies%20have%20shown,Arab%20Report%20for%20Cultural%20Development.(

A recent statistic claimed that on average Arab children read 6 minutes a year according to the Arab Thought Foundation (https://en.unesco.org/creativity/policy-monitoring-platform/arab-reading-challenge#:~:text=Reports%20and%20studies%20have%20shown,Arab%20Report%20for%20Cultural%20Development), compared to their Western counterparts. Hence, we measure the number of minutes spent reading. 

We measure:

• Number of Books read
• Levels Progressed
• Achievement of Learning Outcomes

On our platforms we have divided our content into levels (approximately 2-3 levels per grade). We measure level progression.

Every activity on our platform is aligned with learning outcomes; we measure achievement of learning outcomes.

Theory of Change

Impact

Quality of education for K-12 MENA students that contributes to improved learning outcomes

Long term outcomes

Increased access by students to hundreds of books and videos

Improved students learning outcomes

Longer term impact such as improvement in writing and in all other subjects, as well a developing a habit of reading and learning through reading

Acknowledgement of students achievement

Short term Outcomes

Dissemination of the app to students in K-12 in MENA through schools, governments and NGOs

Students improvement in: literacy skills, reading comprehension and grammar

Students demonstrate positive social behavior and improved confidence

Teachers apply active teaching by identifying individual differences and weakness

Output

A levelled reading gamifed digital library of hundreds of books with related quizzes

Books focused on SDGs and social emotional learning (gender equality, tolerance etc)

Teacher training on using technology in classroom

Parents and teachers involved in in their students education and know areas of weakness

Input

Develop a leveled Arabic literacy/ reading app for students K-12

Source award winning books from leading regional publishers and develop in house more books, videos and assessments

Develop teacher training material

Develop detailed reports

Problem

50% of Children in MENA suffer from learning poverty; they cannot read or write an age appropriate text

Our platform is available to use on both web and mobile devices, the backend is built using PHP CodeIgniter, for the frontend/web we use bootstrap, while for the mobile we have native applications on both Android and iOS, the database is MySQL. All hosted on AWS. 

Below is a detailed description

Software Components:

• IRA: it is a single CodeIgniter installation consisting of 4 applications (Students App, teachers App, Admin interface and the API platform (Same code base).

Platform

technology stack

Web

PHP - OpenSource -

CodeIgniter MVC framework  - OpenSource -

Mysql   - OpenSource -

Firebase

FFmpeg  - OpenSource -

Image Magic  - OpenSource -

supervisord  - OpenSource -

Application database

Mysql  - OpenSource -

Firebase - realtime database by Google

Data warehouse 

Android - Student APP

Java + Kotlin

IOS - Student APP

Swift

Android - Teacher APP

Kotlin

IOS - Teacher APP

Swift

All our applications are hosted on AWS with the following services running:

• EC2: It is used to host web application servers, database servers, monitoring and logging servers

• ELB: Load balancing service, used to distribute users requests among backend application servers

• AWS Autoscaling: Used to auto scale application servers capacity based on requests volume

• EFS: Elastic File service, used to host application static files, content, and file uploads by website users

• S3: Used for file uploads, store backup files for system databases, and application logs archives

• Route53: DNS service used to host websites' DNS zones.

• SES: Simple Email service, used to send application emails to website subscribers, and to send monitoring notification to technical team members.

• AWS Client VPN: Used to create client VPN connections for technical teams to securely access LTM EC2 servers.

•  

Monitoring

• Infrastructure Monitoring

1. Sensu: Sensu is an open source monitoring solution, it's used to monitor servers availability along with servers metrics like CPU, disk and RAM utilization.

1. AWS CloudWatch: It is used to monitor AWS services metrics like healthy application servers, requests count, infrastructure response time and requests errors.

1. Pingdom: it is a global performance and availability monitoring solution for the websites, applications and servers.

• Application Monitoring

1. NewRelic: It's a tool to monitor application response time, collects transaction traces, and reports application level errors and failures.

1. Monyog: It's a monitoring tool for application database servers, it captures database engine queries execution times, slow queries, and internal database metrics like MySQL database engine utilizations.

Logging

Logging in LittleThinkingMinds is based on Filebeat/Logstash system, servers and application logs are shipped by Filebeat daemon in web servers to the logging server, logging servers hosts Logstash service which aggregates the logs and store them in log files, application owners can view aggregated logs in the logging server to analyze system and application warnings and errors.

In Jordan females have very high education rates yet only make up 14% of the workforce.

At LTM we are 65% female. We cater very much to working women through remote working, flexible schedules and a longer maternity leave than the government allows. Our management team is also 60% female.

The Arab world is a very patriarchal society. We are aware of this and as such we make sure our books cover gender equality (female heroes, female role models and males/ boys also partaking in nontraditional tasks such as house chores). We recently also signed a partnership with Rebel Girls, a US multi-platform empowerment brand dedicated to helping raise the most inspired and confident generation of girls. We will be translating 15 of their stories into Arabic and adding them to our library to inspire young girls and even boys. The stories cover women from different races and continents including Africa, US, South America and Middle East.

We also worked on creating 100 new stories with UNICEF that target Syrian refugees; we made sure the books cover different kinds of families (taking into account single mothers, or children raised by extended family) as well as children with disabilities. Social cohesion as per global sustainable development goals were one of our mains drivers in creating the books to ensure the connectedness between Jordanian and Syrian students within the school environment. 

As for our users, we are very passionate that our platforms reach all children in the region. For our projects in the public schools, we have created an offline version of the application for low resource schools that have little access to the internet.

We also offer much lower prices for large government projects (compared to private schools), and we have provided the platform for free for a few remedial education centers that cater to Iraqi and Yemeni refugees in Jordan.

Little Thinking Minds is a subscription based company (SAAS model) where schools pay a yearly subscription to give access to their teachers and students (B2B), and parents pay a yearly subscription for their children (B2C). We also work with governments, NGOs and foundations to provide our tools to marginalized and disadvantaged students.

We currently generate around $2million a year through selling our solution to private schools in the region. Our plan is to scale vertically and horizontally (more schools, more grades and more products) into more private schools as well as penetrate the public-school sector through grants and government funding. 

We have already fundraised $3 million to date. Our last fundraise was in 2019 from regional investors who included

• Algebra Ventures, Egypt: a premier investment partner for building sustainable businesses in Saudi Arabia and the MENA region
• ISSF, Jordan: a private-sector-managed fund promoting the Jordanian entrepreneurial ecosystem by empowering youth-led, early-stage, innovative startups and SMEs.
• Mindshift Capital, UAE- USA: a global, women-run venture firm investing in amazing women-led companies solving big problems with sophisticated solutions.
•  Al Turki Group, Saudi Arabia: a premier investor and partner of choice for building sustainable businesses in Saudi Arabia and the MENA region.

For our work in the public schools we have received funding either through grants (All Children Reading, UNICEF, Shoman Foundation)  or through paid projects and Vodafone Foundation in Egypt.