Through human-AI collaboration, we summarize, link, evaluate, improve, and learn science for long-term.
1Cademy's goal is to streamline the learning process by converting human knowledge into easily digestible micro-content that eliminates redundancy. By combining various perspectives for a given concept within a particular discipline, learners can master this concept by referring to a single comprehensive piece of micro-content. To achieve this, 1Cademy uses human-AI collaboration to gather valuable knowledge from multiple sources and merges it into concise pieces. This collaborative note-taking approach ensures that the micro-content is applicable to various students studying the same topics, beyond the scope of just one semester.
Efficient learning requires progressive learning content without overlapping knowledge that creates redundancy. 1Cademy divides learning content into micro-content that define a single concept and can be used in a variety of learning contexts and goals. This mitigates the redundancy and disorientation of reading a concept defined in different ways from numerous pieces of literature. This can be achieved through a comprehensive "prerequisite knowledge graph" of micro-content. To address competing or conflicting claims on a given concept, we link the differing views to a generic definition of the concept so that the opposed views can be compared in a side-by-side presentation. 1Cademy notebooks are knowledge graphs of these micro-content pieces, shared among many learners and researchers, who continually improve and add perspectives to the content over time.
To ensure the quality of the knowledge graph on 1Cademy, we have implemented an AI-enhanced peer-review process. Each individual concept, represented as a micro-content piece, is evaluated through a collaboration of AI and members of the community, and the score of the piece will determine its possibility for modification or deletion.
We collaborate with each other and AI assistance to improve the micro-content in the knowledge graph by continually updating and refining concepts. For each micro-content piece, there are multiple versions proposed by different people, which cover multiple perspectives and use-cases. 1Cademy visualizes these side-by-side to optimize learning.
1Cademy Assistant uses a personalized question-asking approach, leveraging the prerequisite relations in the 1Cademy knowledge graph as well as the student's prior answers. Upon answering correctly, students are guided towards more advanced topics, while incorrect answers result in a review of the topic's prerequisites to ensure a strong foundation before proceeding. Through this method, students can earn a daily point for answering ten questions correctly, which has been shown through years of experimentation to increase motivation and encourage spaced-out practice throughout the semester. Our research has found that this approach is especially beneficial for students with lower GPAs, ultimately leading to improved exam scores. By combining the power of the 1Cademy knowledge graph, personalized practice history, and the counting days incentive, 1Cademy Assistant is a highly effective tool for facilitating long-term learning.
1Cademy Assistant is an intelligent system that leverages the power of 1Cademy's knowledge graph to provide personalized question answering to learners. By analyzing the learner's history of answers to practice questions and understanding the prerequisite relationships between the topics covered in the course, the system can provide learners with context-specific guidance. This not only helps learners understand what is covered in the course, but also what lies beyond its boundaries. Additionally, 1Cademy Assistant helps learners develop metacognitive skills by providing insights into why they may be struggling with a specific concept. The system can trace these difficulties back to the prerequisite topics they had difficulty with, helping them to build a more comprehensive understanding of the subject. Finally, 1Cademy Assistant provides instructors with insightful reports on each student's personalized learning journey, allowing them to fine-tune their teaching strategies to better serve their students' needs.
Over the past two years, 1543 students and researchers from 183 institutions have participated in a large-scale collaboration effort through 1Cademy. This collaboration has resulted in the creation of 44665 nodes and 235674 prerequisite links between them, which have been proposed through 88167 proposals.
It is truly inspiring to witness the collaborative learning environment that has been fostered at 1Cademy, where students from both top-ranked and low-ranked schools can come together regardless of their background, ethnicity, or socio-economic status.
Through this platform, students are able to share their unique learning pathways, making difficult concepts more accessible to those who may be struggling. The simplified learning pathways offered on 1Cademy have been a valuable resource for those seeking to deepen their understanding of complex subject matter.
Students are able to engage with the content by voting and commenting on the nodes and links that have been created, which not only helps to acknowledge the valuable contributions of their peers but also encourages meaningful collaboration within the learning community.
This approach to learning empowers students to take an active role in their education, making it a more enjoyable and fulfilling experience. By helping one another, students are able to take pride in their contributions to the larger learning community and are able to learn from one another in a truly collaborative way.
Over the past two years joined 1Cademy.
Have participated in a large-scale collaboration effort through 1Cademy
Are generated through this large-scale collaboration.
Are connected between nodes.
The traditional learning approach for students is to start with foundational concepts and gradually work towards more complex topics. On the other hand, researchers often prefer to begin with the advanced topics and work backwards to gain a deeper understanding of the underlying prerequisites. 1Cademy offers an innovative approach to learning by enabling students to emulate the research method and start with advanced topics, then delve into the prerequisites as needed. This approach allows students to learn in a more targeted and efficient manner, similar to the way researchers approach learning.
1Cademy has fostered the development of communities of enthusiasts for various scientific subjects, comprising individuals from diverse educational institutions and research organizations. These enthusiasts share their discoveries and insights on 1Cademy and come together on a weekly basis to delve deeper into their areas of interest. Through these interactions, we gain insight into the cutting-edge research and learning taking place at our collaborators' institutions and are able to draw connections that inspire new research ideas.
The process of meticulously considering the prerequisites for each concept when adding them to 1Cademy not only improves the quality of our learning, but also helps us uncover novel learning pathways to grasp complex concepts that we previously thought were unattainable.
1Cademy members are constantly evaluating the efficacy of the content and learning pathways. If a member discovers a more straightforward method for defining or explaining a concept, they can propose it on 1Cademy for community review. Through this process, the community collectively decides which approach is most effective for learning that particular concept. As a result, the learning experience through 1Cademy continually improves, becoming both more efficient and enjoyable over time.
While information on any topic is readily available on the internet, many people still choose to invest in textbooks and courses. The reason for this is that these resources provide structured learning pathways - step-by-step procedures to achieve one's learning objectives. However, traditional textbooks and courses are limited by the perspectives of a few authors and are infrequently updated or improved. 1Cademy offers a solution to this by providing a collaborative platform for students, instructors, and researchers to design and share learning pathways on any topic, all within the framework of a shared knowledge graph.
Similar to Wikipedia, 1Cademy is built through a collaborative effort on a large scale. However, while Wikipedia is the most comprehensive encyclopedia, 1Cademy's goal is to tap into the collective intelligence of its users to uncover the most efficient learning pathways for any given topic by identifying the most effective prerequisite connections.
Ample research in cognitive psychology has demonstrated that the act of learning with the intention of teaching others is more effective than learning for the sole purpose of being tested. On 1Cademy, we condense and depict our learning pathways with the objective of enhancing the learning experience for our collaborators. In the process, our understanding of the topics deepens as we contemplate ways to make them more accessible for others to learn.
Have you ever encountered difficulty finding relevant content to learn something, because you're not sure what the appropriate keywords are? For instance, what would you search for to learn how to create the web animations featured on a particular website? Simply searching a phrase might not yield the most helpful results. 1Cademy offers a solution to this challenge by providing both a factual search engine and a mechanism for creating a personalized view of the shared knowledge graph to facilitate exploratory search. This way, even without having the exact keywords, one can navigate through the hierarchical structure of concepts and their prerequisite links to facilitate learning.
These days, we see political, sexual, ethnic, or even scientific polarization everywhere on the Internet. Echo chambers are formed where a group of people only accept thoughts and ideas that are aligned with their perspectives, ignoring alternatives views. 1Cademy provides us with a consensus-based collaboration mechanism where alternative or even competing perspectives are placed side-by-side so that one can easily compare and contrast them to learn and rationalize each topic in different contexts.
1Cademy facilitated the formation of communities of learners and researchers who can learn from each other, exchange ideas and support one another in their learning journey.
1Cademy offers a comprehensive and integrated solution that enhances the educational and research experience through its three interconnected systems.
1Cademy is a platform that aims to improve the efficiency of learning and research by utilizing a collaborative approach to gather information from various sources and organize it into concise notes that focus on a single concept.
These notes are granularly organized and visualized as a knowledge graph that illustrates the hierarchical relationships between concepts. The platform uses a peer-review process, reputation system, and voting mechanism to ensure the quality of the knowledge graph and encourage the development of high-quality content.
Through this process, students and researchers can improve upon each other's contributions, propose more up-to-date and user-friendly versions of each note and share their learning perspectives.
Over the past two years, 1,543 students and researchers from 183 institutions have participated in the platform, resulting in the formation of 49 research and learning communities covering a wide range of subjects.
The 1Cademy AI Assistant is designed to improve human life and education by promoting the development of beneficial habits and scheduling tasks and meetings.
The assistant recognizes the positive impact of these habits on one's life and motivates the user to invest more time in them. It auto-schedules tasks and optimizes time-allocation, schedules 1-to-1 and group meetings, and keeps the user in sync with their instructors by providing information on courses, assignment deadlines, classes, and exams.
It also provides real-time updates on the user's progress on tasks and deadlines and rewards them with points and badges. Additionally, it employs techniques such as desirable difficulties and Pomodoro to boost long-term learning and mitigate procrastination and burn-out.
Furthermore, it leverages the psychology of motivation by breaking tasks and habits into small pieces and making losses as prominent as gains to motivate the user to learn from their mistakes.
Introducing 1Cademy Assistant - Practice Tool
Introducing 1Cademy Assistant - Question Answering
Introducing 1Cademy Assistant - Voice-based Practice
School of Information
awarded research credits to host 1Cademy on GCP services, under award number 205607640.
Iman YeckehZaare is the founder and architect of 1Cademy. He is currently pursuing his Ph.D. at the University of Michigan, School of Information. He has a Master of Science Degree in Information Science with two specializations in Human-Computer Interaction (HCI) and Information Economics for Management (IEM) from the same institution. Additionally, Iman holds two Bachelor of Engineering Degrees in Computer Science and Information Technology.
Iman was awarded the title of Best Graduate Student Instructor of the Year 2018-2019 at the University of Michigan, School of Information. He was also a Michigan I-Corps 2013 Graduate, a Campus of the Future 2018 Semi-finalist, an Innovation in Action 2018 2nd Prize awardee, and a Learning Levers 2019 3rd Prize awardee.
Paul Resnick holds the esteemed position of Michael D. Cohen Collegiate Professor of Information, Associate Dean for Research and Innovation, and Professor of Information at the University of Michigan's School of Information. As a trailblazer in the fields of recommender systems and reputation systems, he played a pivotal role in developing the award-winning GroupLens Collaborative Filtering Recommender system, which received the 2010 ACM Software Systems Award.
In recognition of his exceptional work, Resnick received the prestigious University of Michigan Distinguished Faculty Achievement Award in 2016 and the SIGCHI CHI Academy Award in 2017. Among his numerous notable publications, "The Social Cost of Cheap Pseudonyms," co-authored with Eric Friedman, earned the inaugural ACM EC Test of Time Award. Additionally, his 2012 MIT Press book, Building Successful Online Communities: Evidence-based Social Design, co-authored with Robert Kraut, made a significant impact in the field.
In 2020, Resnick was honored as an ACM Fellow for his remarkable contributions to recommender systems, economics and computation, and online communities, an honor reserved for the top one percent of ACM Members. He served as chair of the RecSys Conference steering committee from 2013 to 2015, and in 2014, co-chaired the ICWSM Conference. Resnick obtained his Ph.D. from MIT in 1992. He has been an advisor to the 1Cademy project since 2013.
Joel Podolny, a distinguished sociologist and CEO of Honor Education, Inc., has an impressive background in academia and corporate training. Previously, he held the position of Vice President at Apple and was the founding Dean of Apple University (2009-2021), where he managed the company's internal training program. In addition, he served as Dean and Professor of Management at the Yale School of Management (2005-2008), spearheading a significant overhaul of the Yale MBA curriculum to better equip students for the intricate, cross-functional global landscape.
Prior to his tenure at Yale, Podolny was a Professor of Business Administration and Sociology at Harvard Business School (2002-2005) and a Professor of Organizational Behavior and Strategic Management at Stanford Graduate School of Business (1991-2002). While at Stanford, he held the position of Senior Associate Dean and taught courses in business strategy, organizational behavior, and global management. Podolny earned his Ph.D. in Sociology from Harvard University in 1991.
Roby Harrington is currently a board member of the Camphill Foundation, an advisor to CORE ECON, a board member of governors at Stanford University Press, a special advisor to the CEO of Honor Education Technology, and a farmer at Ten Barn Farm in Ghent, NY. At W. W. Norton & Company, Inc, Roby held various positions, including sales representative (1979-82), editor of political science, philosophy, and religion (1983-2020), national sales manager (1987-93), director of the college department (1994-2020), and Vice Chairman (2007-2021). He was also the chairman of the board at Camphill Foundation (2015-2020) and a fellow at the Center for Advanced Study in the Behavioral Sciences at Stanford University (2020-2021).
Collaborative question generation and mapping by students has been shown to improve students' active (engaged) and meaningful learning. In implementing these methods into a course that enabled high levels of autonomy (as recommended by Self-determination theory), we found student procrastination to be a persistent problem. Since student contribution tends to wane over the course of a semester, the efficacy of a curriculum reliant on content generation similarly wanes. We describe our efforts in reducing procrastination in a course focused on collaborative question generation and mapping, using an iterative design research methodology, over eight months of two semesters. To encourage students to create high-quality questions, we implemented a voting system that graded students based on the number of instructor up-votes given to questions they created. While this reduced procrastination early on, students found ways to take advantage of the autonomy provided by the course curriculum, leading them to resume procrastinating. To address this issue, we adjusted the grading schema for the second semester by scoring both students' and instructors' up-votes, while also allowing students to present their created content in optional weekly meetings. The introduction of optional weekly meetings had the greatest effect towards decreasing procrastination. End-of-semester surveys reported that: 1) while students enjoyed both semesters, they gave more positive feedback for the second semester; 2) student-led discussions for personally created content, when combined with the autonomy to choose both the topic and time to study, helped their learning and time management skills.
Desirable difficulties such as retrieval practice (testing) and spacing (distributed studying) are shown to improve long-term learning. Despite their knowledge about the benefits of retrieval practice, students struggle with application. We propose a mechanism of embedding desirable difficulties in the classroom called "retrieval-based teaching." We define it as asking students many ungraded, granular questions in class. We hypothesized that this method could motivate students to (1) study more and (2) increase the spacing of their studying. We tested these two hypotheses through a quasi-experiment in an introductory programming course. We compared 684 students' granular activities with an interactive eBook between the class discussion sections where the intervention was implemented and the control discussion sections. Over four semesters, there were a total of 17 graduate student instructors (GSIs) that taught the discussion sections. Each semester, there were five discussion sections, each taught by a distinct GSI. Only one of the five per semester implemented the treatment in their discussion section(s) by dedicating most of the class time for retrieval-based teaching. Our analysis of these data collected over four consecutive semesters shows that retrieval-based teaching motivated students to space their studying over an average of 3.78 more days, but it did not significantly increase the amount they studied. Students in the treatment group earned an average of 2.36 percentage points higher in course grades. Our mediation analysis indicates that spacing was the main factor in increasing the treated students' grades.
Prior literature suggests that computer science education (CSE) was less affected by the pandemic than other disciplines. However, it is unclear how the pandemic affected the quality and quantity of students' studying in CSE. We measure the impact of the pandemic on the amount and spacing of students' studying in a large introductory computer science course. Spacing is defined as the distribution of studying over multiple sessions, which is shown to improve long-term learning. Using multiple regression models, we analyzed the total number of students' interactions with the eBook and the number of days they used it, as a proxy for studying amount and spacing, respectively. We compared two sequential winter semesters of the course, one during (Winter 2021) and one prior to the pandemic (Winter 2020). After controlling for possible confounders, the results show that students had 1,345.87 fewer eBook interactions and distributed their studying on 2.36 fewer days during the pandemic when compared to the previous semester prior to the pandemic. We also compared four semesters prior to the pandemic (Fall and Winter of 2018 and 2019) to two semesters during the pandemic (Fall 2020 and Winter 2021). We found, on average, students had 3,376.30 fewer interactions with the eBook and studied the eBook on 16.35 fewer days during the pandemic. Contrary to prior studies, our results indicate that the pandemic negatively affected the amount and spacing of studying in an introductory computer science course, which may have a negative impact on their education.
Spacing and procrastination are often thought of as opposites. It is possible, however, for a student to space their studying by doing something every day throughout the semester and still procrastinate by waiting until late in the semester to increase their amount of studying. To analyze the relationship between spacing and procrastination, we examined 674 students’ interactions with a course eBook over four semesters of an introductory programming course. We measured each student’s semester-level spacing as the number of days they interacted with the eBook, and each student’s semester-level procrastination as the average delay from the start of the semester for all their eBook interactions. Surprisingly, there was a small, yet positive, correlation between the two measures. Which, then, matters for course performance: studying over more days or studying earlier in the semester? When controlling for total amount of studying, as well as a number of academic and demographic characteristics in an SEM analysis, we find a strong positive effect of spacing but no significant effect of procrastination on final exam scores.
Extensive prior research shows that spacing – the distribution of studying over multiple sessions – significantly improves long-term learning in many disciplines. However, in computer science education, it is unclear if 1) spacing is effective in an incentivized, non-imposed setting and 2) when incentivized, female and male students space their studying differently. To investigate these research questions, we examined how students in an introductory computer science course (378 female and 310 male) spaced their studying. A retrieval practice tool in the course (for 5% of the course grade) incentivized students to space their studying, by awarding a point per day of usage. To measure how much each student spaced, we examined their interactions with the course eBook, which served as their primary learning resource. Specifically, when comparing two students with the same academic and demographic characteristics, the same measure of course easiness, and the same amount of content studied, we considered the student who distributed their studying over more days to be the one who spaced more. Using this definition, our structural equation modeling (SEM) results show that, 1) on average, students who spaced their studying over 14.516 more days (one standard deviation) got 2.25% higher final exam scores; and 2) female students spaced their studying over 4.331 more days than their male counterparts. These results suggest that, in an introductory computer science course, incentivized spacing is effective. Notably, when compared to their male counterparts, female students both exhibited more spacing and obtained higher final exam scores through spacing.
Generating multiple-choice questions is known to improve students' critical thinking and deep learning. Visualizing relationships between concepts enhances meaningful learning, students' ability to relate new concepts to previously learned concepts. We designed and deployed a collaborative learning process through which students generate multiple-choice questions and represent the prerequisite knowledge structure between questions as visual links in a shared map, using a variation of Concept Maps that we call "QMap." We conducted a four-month study with 19 undergraduate students. Students sustained voluntary contributions, creating 992 good questions, and drawing 1,255 meaningful links between the questions. Through analyzing self-reports, observations, and usage data, we report on the technical and social design features that led students to sustain their motivation.
Retrieval practice, spacing, and interleaving are known to enhance long-term learning and transfer, but reduce short-term performance. It can be difficult to get both students and instructors to use these techniques since they perceive them as impeding initial student learning. We leveraged user experience design and research techniques, including survey and participant observation, to improve the design of a practice tool during a semester of use in a large introductory Python programming course. In this paper, we describe the design features that made the tool effective for learning as well as motivating. These include requiring spacing by giving credit for each day that a student answered a minimum number of questions, adapting a spaced repetition algorithm to schedule topics rather than specific questions, providing a visual representation of the evolving schedule in order to support meta-cognition, and providing several gameful design elements. To assess effectiveness, we estimated a regression model: each hour spent using the practice tool over the course of a semester was associated with an increase in final exam grades of 1.04%, even after controlling for many potential confounds. To assess motivation, we report on the amount of practice tool use: 62 of the 193 students (32%) voluntarily used the tool more than the required 45 days. This provides evidence that the design of the tool successfully overcame the typically negative perceptions of retrieval practice, spacing, and interleaving.
In an introductory Python programming course intended for non-majors with little prior CS experience, with 85 male and 108 female students, we were able to capture electronic traces of students' studying and problem-solving. There was no significant difference in final exam scores by gender but we found that female students spent 12.1 more hours studying over the semester while male students on average earned 2.7 more points per hour of solving problem set questions over the first half of the semester. We were able to capture their learning behavior because students studied using the Runestone interactive textbook and completed weekly problem sets in the same platform for the first half of the semester. We analyzed these logs to determine three quantities for each student. One is study time, as measured by total use of Runestone outside of weekly assignments. The second is speed, as measured by the number of points students earned per hour working on problem sets. The third is earliness, as measured by how far before the deadlines they worked on weekly assignments. We conclude that male students were faster at completing problem sets early in the semester but that female students found an alternative pathway to success.
The Runestone ebook platform is open source, extensible, and already serves over 25,000 learners a day. The site currently hosts 18 free ebooks for computing courses. Instructors can create a custom course from any of the existing ebooks on the site and can have their students register for that custom course. Instructors can create assignments from the existing material in each ebook, grade assignments, and visualize student progress. Instructors can even create new content for assignments. The Runestone ebooks contain instructional material and a variety of practice problem types with immediate feedback. One of the practice types, Parsons problems, is also adaptive, which means that the difficulty of the problem is based on the learner’s performance. Learner interaction is recorded and can be analyzed. This paper presents the history of Runestone, describes the interactive features, summarizes the previous research studies, and provides detail on the recorded data. Interaction data can be shared with other learning environments through the Learning Tools Interoperability Standard (LTI).
One of the most important aspects of 1Cademy is its unique knowledge representation format. To become a researcher on 1Cademy, you should first engage in one of our ongoing research projects, as a participant. In the project, randomly chosen for you, we will test which type of knowledge representation format works better for your reading comprehension, short-term learning, and long-term learning. This will not only help us improve the design of 1Cademy, but along the way, you will get experience about how to use 1Cademy. For this purpose, you should create an account on our research website and specify your availabilities for three sessions with our UX researchers. In the first session, they will ask you to read two short passages and answer some questions about those passages. This will take an hour. The second and third sessions will be only for 30 minutes each and follow a similar format. Note that it is necessary to complete the second and third sessions, exactly three and seven days after the first session. So, please carefully specify your availability on our research website.