#xAPI Forming Building Blocks of Personal Assistant for Learning
The Experience API (xAPI) provides a means to store and access data about learning experiences. The xAPI is designed to support existing browser-based capabilities such as the Sharable Content Object Reference Model (SCORM) and extend that support to include non-browser-based capabilities such as mobile applications, games, virtual environments, full-scale simulators, and sensors. The xAPI also allows for much more granular data to be collected, such as learner progress, virtual media, and even real-world experiences.
The xAPI introduces the concept of a learning record store (LRS) to create a “learning experience tracking service.” This store is a database of learning records that can be accessed by other systems, such as reporting applications, human resources programs, a learning management system, or any other authorized system.
The xAPI uses a format for the learning record similar to activity streams used in social media. The format can be thought of as a simple sentence with the structure, <actor><verb><object>: “Alex viewed video” or “Melissa read book.” The standard also is meant to accommodate speciﬁc details: “Alex viewed [learning fractions] video” or “Melissa read [The Autobiography of Miss Jane Pittman] book.”
The learning record can infer information based on previously stored data. (A book is read; a video is viewed.) The data contained in those learning records can also be used to build up, assign, or sequence activities for learners: After somebody has written a chapter, for example, another learner can be told to edit it, a third to proof it, and a fourth to add it to an existing chapter. These capabilities allow the xAPI to be used to recommend content, tailor learning experiences, and provide instructor feedback on content or learner assessments.
The xAPI is meant to be “stacked” with other technologies for value-added services. For example, the xAPI can track extremely granular data in several learning record stores, which can then be conﬁgured to sync up with data from other technologies like inBloom or the Open Badges Infrastructure. The xAPI can be used to track usage data of content or sub-components of content. This usage data can then be pushed to the Learning Registry or used by LRMI as valuable paradata for those searching for high-quality educational content. (“Which video on fractions was most assigned by other teachers in my district?”)
The xAPI provides a capability to store and access information from a diverse set of learning, training, and performance support experiences. Teachers can use this data to determine a student’s true understanding of a concept beyond basic summary data such as quiz or test scores. Systems can automatically tailor content based on a thorough understanding of a learner’s previous course history and interactions with learning content. Learners can choose to include informal and self-guided learning experiences into their learning record store that can later be used to help determine the impact of learning outside of formal curricula.
Future plans include the creation of community-speciﬁc vocabularies or proﬁles that will enhance interoperability of content and systems. Sponsoring organization ADL will serve as a catalyst for additional xAPI proﬁles. In addition, xAPI tangential speciﬁcations for big data query interfaces, learner proﬁles, and activity deﬁnitions are all potential extension points.
The xAPI forms a building block for ADL’s eventual goal: to create a “personal assistant for learning” that can anticipate learner needs, integrate with relevant information, and provide access to personalized learning content.
Fox, C., Schaffhauser, D., Fletcher, G., & Levin, D. (2013). Transforming Data to Information in Service of Learning. Washington, DC: State Educational Technology Directors Association (SETDA).