Educational technology can be delivered, utilizing peer-to-peer (P2P) database computing, P2P statistical computing, providing educational data mining (EDM) features while providing students storage of their data on their mobile computers, facilitating data privacy.
Educational technology can interoperate with software at schools’ libraries, public schools or universities, providing transparent configurability to local school districts or local schoolboards. Public school libraries’ software can transparently, configurably, provide semantic content services from a large number of materials, including digital textbooks, services utilized by dialogue systems, spoken dialogue systems or digital characters (e.g. Digital Ira) while interacting with students in scenarios including uses of featureful, interoperable, digital textbooks selected by state or local schoolboards.
Related Standards Activities
IDPF (EPUB), IMS Global (EDUPUB), Digital Publishing Activity, Math Working Group, Multimodal Interaction Activity, Provenance Working Group (Provenance Page at the Semantic Web Wiki), Semantic Web Activity, Speech API Community Group, Voice Browser Working Group
Topical to peer-to-peer (P2P) educational technology systems is that each student’s data, each student’s models, can be stored on their mobile computer while the providing features of educational data mining (EDM). The local storage of student data, of student models, is possible while simultaneously providing EDM features, technological advantages, to students, teachers, administrators and school systems. Numerous architectural models are possible with the premises of: (1) student data privacy, students’ data, students’ models, on students’ mobile computers, (2) interoperability with school library technologies. An architectural concept is described with students’ mobile computers as natural language user interfaces, dialogue systems, spoken dialogue systems or digital characters, to semantic content services provided by schools’ libraries.
With familiarity of users’ contexts, users’ tasks, users’ documents, users’ books, users’ textbooks, software can better perform natural language understanding, semantically interpreting the instantaneous utterances of users. Similarly, software can better perform natural language generation, articulating language to users more featurefully utilizing such data. Similarly, software can better perform handwriting recognition, including of diagrams as well as of scientific or mathematical notations. Systems processing sensor data, e.g. students’ prosody, nonverbal cues or affect, can process such sensor data or derived data locally on students’ mobile computers. Data from multiple software applications, users’ contexts, users’ tasks, users’ documents, users’ books, users’ textbooks, as well as event streams from multiple applications can be processed on users’ machines, complex event processing, data stream mining, derived data stored on users’ machines to provide modeling-related features to multiple software applications on users’ machines. The view presented is that users’ data is reasonably stored on users’ computers.
Interestingly, components on students’ mobile computers, in addition to teachers’ software, can utilize real-time data, including EDM data, from distributed P2P processing. Machine learning heuristics, for example, can be enhanced by harnessing classroom P2P computing, obtaining course-specific contextual data. Students’ mobile computers can, as per P2P computing, process EDM data for students, teachers, administrators or school systems, while students’ data, including increasingly sophisticated student models, are stored on students’ devices. Pertinent to student data privacy topics, both the present as well as the futures of student modeling are important to consider; increasingly advanced modeling components accumulating data, through at least K-12, from uses of software applications, Web browsing, digital textbook uses, per course, through years of students’ learning.
We can envision library system technologies which stream semantic content for dialogue systems, spoken dialogue systems or digital characters, natural language generation, components on mobile computers. Library system technologies can also predict, prepare, content that students’ might request. Students would tend to be simultaneously exploring topics pertaining to or related to topics in digital textbooks or curriculum. School library systems can utilize data from materials including the textbooks, syllabi and websites of students’ courses.
We can envision content on the sides of students’ screens, hyperlinks to materials or into digital textbooks, concurrent to digital tutors’ utterances while digital tutors, dialogue systems, spoken dialogue systems or digital characters, present topics to students or answer students’ questions. Such content could include the provenance of the semantic content processed into utterances by dialogue systems, spoken dialogue systems or digital characters. Semantic data, streamed from school libraries, articulated into natural language utilizing students’ data on students’ mobile computers, could have provenances including from the processed content of numerous resources and materials.
In addition to the provenances of semantic content, from resources and materials at school libraries or websites processed by schools’ library systems, the content on the sides of students’ screens, concurrent to the utterances of dialogue systems, spoken dialogues or digital characters, could include hyperlinks to recommended content which relates to the topics uttered.
To estimate when students’ mobile computing hardware will be ready to compute advanced dialogue systems, concurrent to students’ tasks, scientists can utilize data including graphics card manufacturers’ roadmaps, where such projections include advancements including stacked DRAM. Technical topics include mobile computing batteries, graphics chips’ processing throughputs, graphics chips’ electrical efficiencies as well as the number of such chips in mobile computers. At CES 2014, Digital Ira was presented on a mobile device.
Advantages of EDM are possible for students, teachers, administrators and school systems without any storage, collection, harvesting, processing, productization or utilization of students’ data by third parties. Advantages of EDM are possible for students, teachers, administrators and school systems while students’ data is stored on students’ mobile computers. Advancements to databases (http://lists.w3.org/Archives/Public/www-math/2014May/0008.html) and to digital textbooks (http://www.w3.org/community/argumentation/2014/08/12/document-personalization-and-user-data-privacy-client-side-document-processing-utilizing-locally-stored-user-data-and-user-models/) were indicated.
Distributed computation, decentralized computation, P2P computation, utilizing students’ mobile computers, including in public school settings, increases the number of EDM features which can be provided to students, teachers, administrators or school systems, while students’ data, including increasingly sophisticated student models, can be stored locally on their mobile computers, protecting students’ data privacy. The storage of students’ various data, including student models, other than upon their mobile computers is not necessary to provide students, teachers, administrators or schools systems with EDM features.
Library technologies can provide semantic content services, knowledgebase services, transparently configurable by school districts, interoperating with dialogue systems, spoken dialogue systems or digital characters, digital tutoring technologies, on students’ mobile computers.
 Adomavicius, Gediminas, and Alexander Tuzhilin. “Context-aware recommender systems.” In Recommender systems handbook, pp. 217-253. Springer US, 2011.
 Cugola, Gianpaolo, and Alessandro Margara. “Processing flows of information: From data stream to complex event processing.” ACM Computing Surveys (CSUR) 44, no. 3 (2012): 15.
 Cugola, Gianpaolo, and Alessandro Margara. “Low latency complex event processing on parallel hardware.” Journal of Parallel and Distributed Computing 72, no. 2 (2012): 205-218.
 Koychev, Ivan, and Ingo Schwab. “Adaptation to drifting user’s interests.” InProceedings of ECML2000 Workshop: Machine Learning in New Information Age, pp. 39-46. 2000.
 Kuflik, Vania Dimitrova Tsvi, David Chin Francesco Ricci, and Peter Dolog Geert-Jan Houben. “User Modeling, Adaptation, and Personalization.”
 Luckham, David C., and Brian Frasca. “Complex event processing in distributed systems.” Computer Systems Laboratory Technical Report CSL-TR-98-754. Stanford University, Stanford 28 (1998).
 Özsu, M. Tamer, and Patrick Valduriez. Principles of distributed database systems. Springer, 2011.
 Patel Kuldip, L., M. Savalia Jay, and Sreevalsan-Nair Jaya. “Parallelization of Complex Event Processing on GPU.” HiPC in (2011).
 Resnick, Paul, and Hal R. Varian. “Recommender systems.” Communications of the ACM 40, no. 3 (1997): 56-58.
 Schmidberger, Markus, Martin Morgan, Dirk Eddelbuettel, Hao Yu, Luke Tierney, and Ulrich Mansmann. “State-of-the-art in Parallel Computing with R.” Journal of Statistical Software 47, no. 1 (2009).
 Suchard, Marc A., Quanli Wang, Cliburn Chan, Jacob Frelinger, Andrew Cron, and Mike West. “Understanding GPU programming for statistical computation: Studies in massively parallel massive mixtures.” Journal of Computational and Graphical Statistics 19, no. 2 (2010): 419-438.
 Tanenbaum, Andrew, and Maarten Van Steen. Distributed systems. Pearson Prentice Hall, 2007.
 Turaga, Pavan, Rama Chellappa, Venkatramana S. Subrahmanian, and Octavian Udrea. “Machine recognition of human activities: A survey.” Circuits and Systems for Video Technology, IEEE Transactions on 18, no. 11 (2008): 1473-1488.
 VanLehn, Kurt. “Student modeling.” Foundations of intelligent tutoring systems 55 (1988): 78.
 Watt, Michael G. “Research on the Textbook Selection Process in the United States of America.” Online Submission (2009).
 Webb, Geoffrey I., Michael J. Pazzani, and Daniel Billsus. “Machine learning for user modeling.” User modeling and user-adapted interaction 11, no. 1-2 (2001): 19-29.
 Wegener, Dennis, Thierry Sengstag, Stelios Sfakianakis, Stefan Rüping, and Anthony Assi. “GridR: An R-based tool for scientific data analysis in grid environments.” Future Generation Computer Systems 25, no. 4 (2009): 481-488.
 Wu, Eugene, Yanlei Diao, and Shariq Rizvi. “High-performance complex event processing over streams.” In Proceedings of the 2006 ACM SIGMOD international conference on Management of data, pp. 407-418. ACM, 2006.
 Yoginath, Srikanth B., Nagiza F. Samatova, David Bauer, Guruprasad Kora, George Fann, and Al Geist. “RScaLAPACK: High-Performance Parallel Statistical Computing with R and ScaLAPACK.” In ISCA PDCS, pp. 61-67. 2005.