Cognitive Computing in Education
In 2011, IBM’s Watson computer system beat the top Jeopardy champions and demonstrated that computers could be programmed to understand natural language and make swift and accurate conclusions from large amounts of unstructured data. Recently, a program designed by DeepMind, beat the world champion of the very complex game known as Go.
While these achievements in technology have been first witnessed in the context of gaming, they will eventually make their way into applications that have more practical and economic value. Education is a field that is primed and waiting for computing power of this type to help it solve some of its own unique problems.
Education and the Unstructured Data Problem
In order for information technology to be most effective, it needs access to information. Watson and DeepMind offer solutions to the field of education in dealing with the “unstructured” nature of the information used and produced by education. As much as eighty percent of all the world’s data remains in an “unstructured” condition. Unstructured data is classified as such because it is data that “either does not have a pre-defined data model or is not organized in a pre-defined manner. It is typically text-heavy, but may contain data such as dates, numbers, and facts as well. Unstructured data is all of the information that remains difficult to search by current computer algorithms; either because the information is not in a “search friendly” format or the information resides off the internet in libraries or in offline computers, servers, or hard copy files.
Unstructured data is a particular problem to education because the sort of information that educators use as the “raw material” is the sort of data that is not easily categorized or searchable. For example, teacher lesson plans, and the student work that results from them, are typically not stored in cross-referenced, fielded, and tagged databases. Educational journals may be tagged online for easy searching, but when it comes to wading through the volume of information and make sense of the research contained therein, one comes to understand how daunting the problem of making meaning from unstructured data really is.
Cognitive Computing
Cognitive computing platforms may have an answer to the problems associated with unstructured data. “Cognitive computing refers to systems that learn at scale, reason with purpose and interact with humans naturally. Rather than being explicitly programmed, they learn and reason from their interactions with us and from their experiences with their environment” (Kelly, 2015).
Until IBM’s Watson, computer programs conducted searches in unstructured documents primarily based on searching through data for keywords and phrases. Cognitive computing platforms such as Watson can not only read texts, but it is able to derive meaning from from the text. For the first time in history, Watson demonstrated that a program could be written that had the ability to not only understand natural language, but to make accurate predictions and inferences based solely on information it gained from “reading” unstructured text.
Why Education Needs Cognitive Computing
The purpose of education in today’s world is to equip students to survive and thrive in a fast-paced, ever changing, global information society. With those ends in mind, harnessing information and the technology that facilitates it, is of key importance in the field of education. With the exponential growth of knowledge, and the realization that the vast majority of that knowledge is unstructured, it is critical that the field of education turn its attention to new methods of accessing and analyzing that information.
With the power of cognitive computing systems, educators will be able to utilize the vast resources of unstructured data that they have at their disposal to help answer questions that they have never been able to answer before. Cognitive computing provides a method by which machine learning and “big data” can begin to make important discoveries in the field of education as it has already begun to have in others.
Cognitive Computing and Education
IBM’s Watson is already working with oncologists at leading cancer treatment hospitals to churn through the mountains of cancer research data and suggest treatment plans for patients based upon their individual situations. It is not farfetched to imagine a cognitive computing program that has access to student data in the form of grades, attendance, test scores, and even individual work samples. The program will analyze these datasets, compare them with other similar students, read through the existing database of studies on best practice, and then find the right lesson, video, or teaching technique to suggest to the teacher for that particular student. The program can take this knowledge and prescribe a detailed plan for each student. It can be used by teachers to predict which students are going to have a difficult time with a particular concept and then provide them with the resources that the teacher would need to help the student.
Cognitive Computing and Special Education
Within the field of education, the area where cognitive computing has the potential to become a powerful tool on behalf of the students is in the area of special education. The numbers of students in the United States who qualify for special education services continues to rise each year, and the number of teachers who teach special education and those who are entering the field are also declining (Sawchuck, 2014). This reality is placing a great strain on the system in terms of the lack of personnel and the lack of adequate professional knowledge to support this group of students. Cognitive computing can help provide the help and expertise that is needed to bridge that gap.
By virtue of its development and focus around Individual Education Plans (IEPs), special education that has the largest amount of unstructured data as a result of the copious evaluations, assessments, and monitoring that is required to take place for all special education students. This data is taken regularly on every student as forms a consistent core of data that is collected on every student in the program. The program could also look at the student’s profile, compare it to the research database and the outcomes of other similar students; then recommend a plan of action for the individual student. Cognitive computing addresses all of the fundamental aspects of special education by providing powerful insight into the unstructured data which will lead to more effective individualized education plans for the most needy students.
Conclusions and Recommendations
Cognitive computing platforms are poised to make a major impact in the field of education due to the field’s large amount of (and reliance on) unstructured data. IBM’s Watson has demonstrated that cognitive computing platforms can understand natural language, make predictions, and draw meaningful insight from a wide variety of unstructured data sources. Cognitive computing creates a pathway for educators to search the existing research data, and combine it with student data to draw important conclusions, and to make individualized recommendations. Educators and computer scientists should begin to explore this potential partnership and work together to research and create new tools based on the example of IBM’s success with Watson. Together, they will expand the knowledge of the field of education, and provide great help to the most academically needy students.
References
Sawchuck, S. (2014). Steep drops seen in teacher-prep enrollment numbers, Education Week 34(9), retrieved from: http://www.edweek.org/ew/articles/2014/10/22/09enroll.h34.html
.
