University of Phoenix
Using Artificial Augmentation Intelligence for the Next Generation of Scholarly Writing Tools Dale Crowe, University of Phoenix School of Advanced Studiesdalecrowe@email.phoenix.eduMartin LaPierre, University of Phoenix School of Advanced Studiesmlapierre@email.phoenix.edu AbstractInstructional designers and programmers recognize that Artificial Augmentation Intelligence (AAI) is moving beyond the realm of science fiction and is becoming a reality. An AAI knowledge based application is currently being used in the United States health care system as well as the business sector. Given the nature of AAI it is plausible to use what has been learned in health care and business in educational tool based applications. The primary focus of the qualitative exploratory single case study was to take one aspect of existing academic tools (scholarly writing) and examine the strengths and weaknesses of each with the end result of developing an AAI based prototype Scholarly Writing Software (SWS) application based on IBM Watson’s architecture. Results from analysis of 20 subject mater expert participants demonstrated that AAI has the potential to interact with a word processing program (e.g. Microsoft Word?) to provide real time suggestions and recommendations to improve scholarly writing, including syntactic and semantic recognition. It is possible to program several of the writing style rules (APA, Chicago, MLA, etc.), into an AAI scholarly writing program. Success of the SWS prototype may lead to additional AAI computer/human interaction programs for both K-12 and higher education. Keywords Artificial Augmentation Intelligence in Education – Cognitive Computing- Knowledge Based Systems for Educational ToolsIntroductionInstructional Designers/Technologists, in partnership with information technology researchers, are at the cusp of moving into the next level of technology enhancement, knowledge based applications. Corporations including IBM with Watson (Kelly, 2015); Microsoft Knowledge Base and MS MARCO (Wheatley, 2016); Google Knowledge Vault/Now and SyntaxNet (Slav, 2016); are beginning to offer, to a limited extent open source, access to instructional designers, programmers, and subject matter experts, the inherent capabilities to create applications and instructional models with Natural Language Processing (NLP) abilities. NLP is the ability of a computer program to understand human speech as it is spoken, and is a component of Artificial Augmentation Intelligence (AAI). Other companies like Facebook and Amazon are aggressively moving forward in the development of knowledge based AAI systems (Weinberger, 2016). Outside of the United States “Deep Mind”, located in London, UK, is considered a global leader in AAI innovation (Shead, 2016). According to Akerkar and Sajja (2010), an AAI knowledge based system is, a method that uses artificial intelligence (AI) to solve problems. It houses a depository in the form of database(s) of expert knowledge with functions designed to facilitate the knowledge retrieval in response to specific queries coupled with learning and validation (p.1). An AAI application can offer modules that can analyze visual content so that graphics, tables, and text can be analyzed. Modules offered by the use of AAI provide the necessary tools required to build a word processor hosted (e.g. Microsoft Word?) AAI Scholarly Writing Software (SWS) application with the inherent abilities of computer/human interaction. An SWS knowledge-based application can be programmed to accommodate several different recognized scholarly writing styles (e.g. APA, Modern Language Association (MLA), Chicago, etc.). It is not beyond the realm of possibility that an SWS application can be developed that provides constant computer/human interaction feedback and assessment in real time. When developed, the system is programed to understand what the student is writing and provide feedback in a simultaneous fashion. This may not only help the student learn but also provides the student with the sense that they become more prepared instructionally. Students who are better prepared and complete their course work tend to stay in academic programs longer (Varol & Varol, 2014). Faculty may benefit from a scholarly writing program not only in assisting students, but also as a tool for writing and publishing their own research.Knowledge based systems is not something referred to in instructional technology and information technology circles as “vapor ware”. The term is generally accepted to mean, “a computer-related product that has been widely advertised but has not and may never become available” (Webster’s, 2015). Applications using AAI are being used in hospitals to input information into patient electronic medical records directly from the information doctors dictate into their smart phones. These systems have the intelligence to understand what the medical doctor, physician’s assistant, or nurse is communicating and can correctly place the information into the correct place in the patient medical record (Giles & Wilcox, 2015). Hospital billing departments are using AAI to parse out billable actions that doctors may take on behalf of the patient for a more complete and timely cost accounting system (Giles & Wilcox, 2015). Wagle (2013) premised that the use of AAI in healthcare can be ported over from health care and adapted other areas. Problem, Purpose and SignificanceOnline students (both higher education and K-12) have special needs that make traditional classroom based learning and corresponding policies somewhat difficult. Education administrators and faculty are aware of these issues and are challenged to incorporate enhanced flexible support tools for student and faculty. Online/distance learning has its own unique challenges including, but not limited to, social isolation, the lack of face-to-face academic instruction, and teacher/faculty office hours. Online courses that provide instructional/knowledge based AAI simulation experiences can be especially helpful for both the student as well as the faculty member. Development of AAI based scholarly writing software may be beneficial to traditional ground based, K-12, and higher education institutions as well. Educators at times experience difficulty in instructing students to the importance of addressing topics directly in the online environment (Jones, 2014). Aligning a response with a topic may seem like a simplistic and straightforward proposition for a student to accomplish. When writing scholarly papers students may be of the opinion that long and rambling response will somehow serve to address the topic. Some students do this so regularly that it is difficult for faculty members to address each instance individually and provide proper guidance on what is acceptable when writing a scholarly research paper (Jones, 2014). In addition, students may not even be following all or part of the instructions for the assignments. Incorporating an AAI knowledge based SWS application may make it possible to extend the capabilities of a word processor far beyond what is available now, even with current grammar tools including, Grammarly, White Smoke, Correct English Complete. Knowledge based AAI software may provide enhanced computer/human assistance in the art of scholarly writing more directly and completely that can in turn promote learning. Utilizing a scholarly writing software that provides constant feedback and assessment would not only help the student learn but may also provide the student with the sense that they are better prepared and can complete their course of instruction successfully. Students who are prepared and complete their course work tend to stay in academic programs longer than those who are not (Varol & Varol, 2014). The first phase of the research study was to conduct a literature and critical review of current academic writing tools presently in use, examining and reporting on both their strengths and weaknesses and a determination of/for a potential conceptual framework. The second phase consisted of a development plan that resulted in a prototype SWS application utilizing Watson’s API cloud services. It should be clear that the purpose of this study was not to develop the actual software for distribution. This was beyond the scope of the research study. Literature ReviewAccording to Aytekin, AbdulAziz, Barakat, and Abdurrahman (2012) instructional design research has focused historically on the increase of learner competencies by examining areas of cognitive load theory, and what instructional designers do traditionally to focus on increasing learner efficiencies. Artificial Intelligence simulations are tools that can be used to enhance instructional efficiency for students to learn hands on skills. These include the art of scholarly writing, languages, mathematics, statistics, physical sciences, humanities and other disciplines. In the example of scholarly writing, a problem with most higher education institutions offering both brick and mortar/online curriculum is that the student is expected to have mastered the required writing skills they need before they even begin introductory academic courses (Varol & Varol, 2014). Students experience this problem to some degree in most every discipline taught and it exists in many colleges and universities (Kellogg & Raulerson, 2007). At times it may be a matter of chance finding the right instructor who can teach required skills in a timely manner so as not to let students fall behind their peers academically. Those that do fall behind tend to drop out, try again at a later date or never attempt to take that particular course of instruction again (Kellogg & Raulerson, 2007). One aspect of learning incorporates the styles, attitudes, and approaches that each generation of learners acquires. As an indirect result of the Information Age new classifications have emerged that in effect classify people into age groups. While there are other classifications, the most prevalent that encompass college students and upcoming students, and faculty are, Boomers (1946-1964); Generation X (1965-1980); Generation Y/Millennium (1981-2000) and Generation Z/Boomlets (Born after 2000). According to Oblinger (2003), there exists an aging infrastructure at ground-based universities. In addition the, “…lecture tradition of colleges and universities may not meet the expectations of students raised on the Internet and video games” (p. 44). While this article is somewhat dated (2003) the premise is still relevant in 2017. Students expect to receive experiential, interactive, and authentic learning. This includes the next generation of software tools. To build on Oblinger’s article (2003) came the rapid rise of social networking (Facebook, Twitter, You Tube, Hash Tags, etc.), and there is possibly more on the horizon (e.g. Apple Watch released in April, 2015 and in 2016 Apple Watch 2). Society now expects instantaneous and interactive technologies. With the exponential growth of online learning came new social concerns. According to Minocha (2009, p. 355), there is a motivation for students to become less isolated and embrace technologies in the form of network tools that support and encourage interaction in “time, space, and presence”. While not a substitute for human beings, AAI knowledge-based systems do have the capability to take a step towards the goal of human like interaction. The potential of creating a scholarly writing software that extends the abilities of the common word processor may be a practicality using instructional design/technology principles, natural language processing and free form text organization found in knowledge based systems (Ferrucci, et al., 2010). IBM’s Watson (Watson) is a good example of how robust a knowledge-based system can be utilized. Watson’s knowledge based system bested a pair of human champions on the TV game show Jeopardy in 2011(Watson wins on jeopardy, 2011). Using AAI Watson’s was able to win because it is able to zero in on key words and then can search its memory for clusters of associations that it rigorously checks the top “hits” against all the contextual information the knowledge base system can assemble including, category name; the kind of answer being sought; the time, place, and gender hinted at in the clue; and so on. And when it believes statistically "sure" enough that it has the correct answer it decides to hypothetically hit the buzzer (Anon, 2011). Watson had all the rules for Jeopardy programed into it. By the use of limited “Open Access” and other tools (e.g. Microsoft MS MARCO ) it may be possible to program the rules of scholarly writing into a word processing application that accesses the backend processes that Watson (or similar system) used to win Jeopardy. Corporations like IBM, and others, have built Application Programming Interfaces (API’s) that allow for software to be written that can utilize the AAI knowledge based system capabilities by connecting via the Internet to “Cloud” based servers. (Barinka, 2013). It should be noted that IBM is actively promoting this technology for education (IBM Education, 2016). The AAI backend would not only be able to assess how well the student is addressing a particular topic but it could also check the veracity of declarative sentences in real time. If the student is found to be incorrect the AAI can make suggestions on references that discuss the topic under consideration by the student through instant feedback. For example, the IBM Watson platform can read 200 million pages of text in 3 seconds (Watson Health, 2016). It is possible that the AAI may be able to suggest more appropriate uses of “style” (APA, MLA, etc.) by the student as well. Incorporation of style sets/rules may be feasible by using Microsoft Word? as the interface, requires a language, which works closely with Net Framework. One programming language option would be C Sharp as this works directly with Microsoft Word. Conceptual FrameworkAccording to Miles and Huberman (1994, p.440), conceptual framework “lays out the key factors, constructs, or variables, and presumes relationships among them”. Rather than offering a theoretical explanation, as do quantitative models, conceptual frameworks provide understanding in which conceptual frameworks can be developed and constructed through a process of qualitative analysis (Jabareen, 2009). Physical technology devices that instructors and students use to access their instructional environment will continue to become more prevalent and feature-rich as new products are released into the market. Technology can be used for quality learning, teaching, and assessment. These lead to various interactions that can be summarized as cognitive, affective, and managerial/administrative (Coppola et al., 2002). According to Keengwe and Kidd (2010), one of the first things that must be determined is to assess the [virtual/physical] space that will determine the educational experience. Educational activities in a technology rich environment provide both new opportunities and responsibilities/accountabilities for students, instructors, and the education institution in general. Effective use of technology should include, at minimum, a strategic approach to the design and developing instructional/tool based software that not only suggests corrections but centers on the learning activities and assessment tasks. According to Ablin (2008), the incorporation of a plan to reprocess and repurpose various materials and curriculum designs, in addition to using sound pedagogical strategies is how students learn. This is instrumental for a student’s successful experience. Bernard et al. (2009) conducted a meta-analysis of distance and online learning. Their results quantified the importance of three types of interaction: among students, between the instructor and students and course content. The major conclusion was that the design of interaction technologies into online course has a positive impact on student learning. Bernard et al believed that the first generation of interactive technologies has been limited. Arbrami et al. (2011) went on to “highlight” several evidenced based approaches for the next generation of online learning. Included are, principles and applications arising from theories of “self-regulation, multimedia learning, research based motivational principles, and collaborative learning principles” (p. 53). Theories of self-regulation and multimedia learning coupled collaborative learning principles are used as a foundation of AAI. Arbrami et al. (2011) further posited that later generations of distance education solutions, including but not limited to, two-way video, audio conferencing, and Web based courses could be based on synchronous platforms. AAI would fit within this requirement. Finally, the AAI scholarly writing software proposed is a knowledge tool. A knowledge tool is the next generation of software frameworks that supports student learning coupled with the premise that instructional technology/instructional design is, “… designed specifically to promote student self-regulation in blended, online and distance learning contexts” (p. 90).METHODSResearch QuestionsThe purpose of the qualitative exploratory single case study was to conduct a comprehensive evaluation and critical review of existing academic writing software tools investigating strengths and weaknesses of each; coupled with the practicality of migrating these existing tools to an AAI knowledge-based platform. A second goal is whether a conceptual framework exists in current academic scholarly writing programs that may be incorporated into a development plan/model that can then be migrated to a prototype and future AAI scholarly writing solution. Research questions were:RQ1 What is the strengths and weaknesses of current academic writing software?RQ2 What are the contributing factors for developing a successful AI knowledge-based scholarly writing software?RQ3 What are some unique challenges facing instructional designers and information technology developers in producing knowledge-based scholarly writing software? RQ4 What is the required components (technical, conceptual, etc.) for a knowledge based application that can identify syntactic and semantic recognition?RQ5 What is a practical plan for developing a scholarly writing software program?Design FrameworkAfter determining in the initial exploration that current academic scholarly writing tools and/or knowledge-based systems embody conceptual framework(s) using Jabareen’s (2009) procedure for planned critical review was deemed to be relevant. Steps (“phases”) in the qualitative critical review and development plan/model were used including:Design Framework Phase 1: Mapping the Selected Data Sources.Design Framework Phase 2: Extensive Reading and Categorizing of the Selected Data. Design Framework Phase 3: Identifying and Naming Concepts.Design Framework Phase 4: Deconstructing and Categorizing the Concepts.Design Framework Phase 5: Integrating Concepts/Themes/Categories. Design Framework Phase 6: Synthesis, Resynsthesis, and Making Sense of the Results.Design Framework Phase 7: Validating the Conceptual Framework. Successful development of SWS may lead to other knowledge based systems being developed for not only higher education but K-12 education as well. Potential Differences Between a Potential Knowledge-Based Scholarly Writing Software and Current APA and Grammar Checking Software A comprehensive review of existing scholarly writing tools was conducted including Grammarly, Turnitin, White Smoke, RFW, Perrla, Format Ease, and others. Primary differences between the software packages reviewed and what knowledge-based/AAI software may accomplish is that it augments Microsoft Word? (or another word processor program) and then evaluates the content of what was written both semantically and syntactically. AAI would able to recognize if a user answered a question as required or addressed the topic under discussion in a meaningful manner. It may be also be able to tell if the text was aligned throughout the document, followed directions, and checked for content. The software may be able to check for patterns in writing and content. Another benefit is that a knowledge-based scholarly writing application may provide guidance on alignment and style. It may make suggestions about altering content to match citations or to change citations all together. It can possibly alert the user if they answered a question completely or only partially. An AAI knowledge based scholarly writing program should be able to discern whether the question(s) were answered and respond to citation issues in a limited (scripted) fashion. While some of the software applications reviewed may give the appearance of containing full AI capabilities they fall short of the next iteration of computer/human interface/personality traits knowledge-based software. Population, Sampling, Data Collection & AnalysisPopulation & Sampling. The population for this single exploratory case study consisted of 20 software developers/engineers/instructional designers familiar with traditional and knowledge based software. Snowball sampling in conjunction with purposeful sampling was used for this study. The primary reason for the use of these sampling techniques is the limited number of experts available to the researchers in the area of AAI/AI.Data Collection. A pilot study was conducted with 5 participants The pilot test assisted in determining potential flaws, ambiguity, weakness, limitations, etc. within the research design and interview questions (Turner, 2010). After conducting the pilot study several interview questions were revised. Twenty participants responded to 13 semi structured interview questions developed for this study.Data Analysis. Data analysis consisted of documentation from the individual interviews, journal articles, white papers, and corporate/developer Web sites as part of the triangulation process. In addition, participants were asked to member check their responses. For the individual interviews an inductive approach the data was then analyzed for categories, themes and similarities. NVivo Version 11 software was used to transcribe the individual interviews, field notes, and member checking that were manipulated and analyzed. NVivo software was used to assist in identifying categories representing common themes that participants expressed.Horizontalization. Initial interviews were transcribed into MS Word documents using NVivo 11. Interview documents were exported into an Excel spreadsheet and grouped with their corresponding research question to further engage and deconstruct the transcripts. Individual transcripts were deconstructed into statements that were carefully reviewed. Common categories were then created to develop the clusters, followed by a query through NVivo 11, to establish preliminary units of meaning or themes. Textural and structural descriptions further describe the participants’ knowledge and expertise in the areas of AAI systems and whether the technology is sufficiently viable to produce, at minimum, a prototype scholarly writing software program that incorporates both programming and instructional design characteristics. Several interview questions were consolidated in order to construct more accurate themes. Individual interview and cluster statements and meaning units emerged from the data and are reported in the following tables.Table SEQ Table \* ARABIC 1 Individual Respondents’ Strengths and weaknesses of current academic writing software Statement ClustersExploratory Meaning“Current Academic writing software only helps with some aspects of scholarly writing. There is no full spectrum package for scholarly writing.”Understands the current capabilities and limitations of current software writing tools. “The software is not built to teach but instead only corrects certain mistakes.”Does not provide in depth assistance in an interactive way.“Academic writing software does help with correcting certain aspects of scholarly writing but does not teach you how to actually perform the task of scholarly writing.” Does not provide knowledge based assistance (Artificial Intelligence/Augmentation) Table SEQ Table \* ARABIC 2 Individual Respondents’ Word processors and text editors that address the issue of prolixity (Prolixity - excessive?wordiness?in?speech?or?writing)Statement ClustersExploratory Meaning“Besides using some sort of word count, prolixity is beyond the abilities of most computers.”Present thinking precludes current software programs with addressing prolixity.“Only humans can deal with the issue of prolixity.”Technology cannot solve the issue of prolixity. “I don’t know that there is actually a software package that helps with prolixity.”Unknown whether that there is an existing software program that deals with prolixity.Table SEQ Table \* ARABIC 3 Individual Respondents’ Contributing factors for developing successful AAI knowledge-based scholarly writing software Statement ClustersExploratory Meaning“A professional team of software engineers, instructional designers, and subject matter experts needs to understand all aspects of scholarly writing, computer programing and AI [AAI] approaches to writing software.”Team effort. Expertise by professionals in several different fields involved in the development process. “You would need a team of professionals who understand both AI and scholarly writing.”Experts versed in both AAI/AI and scholarly writing.Table SEQ Table \* ARABIC 4 Individual Respondents’ Unique challenges facing instructional designers and information technology developers in producing knowledge-based scholarly writing softwareStatement ClustersExploratory Meaning“English language, both the written and spoken word, is too complex for computers to understand as humans understand it.”Other than emerging technologies (e.g. IBM Watson) the English language is too complex for existing computers “The English language does not exactly follow computer programming rules.”Due to the complexity of the English language current software writing tools are unable to go beyond basic grammar, mechanics and style.Table SEQ Table \* ARABIC 5 Individual Respondents’ Required components (technical, conceptual, etc.) for a knowledge-based application that can perform syntactic and semantic recognition Commercially available artificial intelligence system that runs on a distributed computing platformStatement ClustersExploratory Meaning“The components of a knowledge based application that can perform syntactic and semantic recognition are the same as those of an expert system such as a knowledge base, Inference engine and a sophisticated user interface capable of advanced speech and text recognition.”Understands the requirements to develop a prototype.“The computer that can perform syntactic and semantic recognition is one that would require advanced Neural Net hardware”Nineteen of the participants believe an AAI system is Feasible.“Probably IBM’s Watson, Kurzweil’s AI tools or Google’s AI tools has all the needed components to build a program that can act as an artificial scholarly writing instructor.”Several applications, other than Watson could be used to develop scholarly writing software.Artificial Augmentation Knowledge-Based Scholarly Writing Software Application Development PrototypeScholarly Writing Software – OverviewScholarly Writing Software (SWS) is a prototype program that works inside of the word processing (e.g. Microsoft Word?) application. The purpose is to aid a user with the task of scholarly writing, and the concept is to go beyond the current abilities of the Microsoft’s Word application and other software writing add-ons. A potential goal of the SWS is to analyze the content of what the individual is writing so that suggestions about improving the text can be made and acted upon in real time. Recent breakthroughs in symbolic and sub-symbolic AAI are encapsulated in subscription based services make understanding a document on the semantic and syntactic level possible. A hypothetical SWS may consist of three main components: User portal – Institution; management servers, and/or commercial AAI processing engine cloud (Figure 1). Figure 1. Scholarly Writing Software Main Components .Institution Management ServersInstitution Management Servers (IMS) are a group of servers that control access to the functions of the SWS from the User Portal (UP). The IMS will contain user identity management services, style knowledge bases and all records of user queries and other usage data. It may be possible for APA style (or other styles) to be programmed into a knowledge base that can be added to the corrections and issues that Microsoft Word? can manage and correct. IMS can be hosted on a dedicated group of servers or on infrastructure as a service cloud computer. IMS can keep track of all costs associated with user queries to the Commercial AAI Processing Engine Cloud, see Figure 2. IMS is the core of the SWS system because it controls all aspects of system usage. UP software is updated, tested on the IMS and users down load the Microsoft Word add-ons form the IMS servers. Figure 2. Institution Management Server cloud based artificial augmentation solution. IMS may provide metrics that instructors can use to evaluate the progress of their students. IMS can also be used to compare the progress of all who use the SWS in an unbiased and impartial manner, allowing for direct comparisons of student progress in scholarly writing. Potential unlimited scalability of the IMS means that there may be technical limits to the number of users that could potentially be using the system mercial AAI Processing Engine CloudCommercial AAI Processing Engine Cloud (CAPEC) is a reference to a fee for service engine (e.g. IBM Watson). It is anticipated that more developers, due to costs, will use CAPEC. For the purposes of the SWS prototype, Watson’s API (CAPWC) was used rather than the costly alternative of institution management servers. It was also decided to use cloud based services to avoid intellectual property concerns. One tool utilized for the prototype was IBM Watson Discovery?. This “service” provides the possibility to quickly construct cognitive, cloud based exploration applications. Using “Discovery” provides insights to unstructured (dark) data. Further, data can come from numerous sources including proprietary, public, and third party (IBM Developer, 2017). Figure 3 provides a model of SWS prototype development.Figure 3. Scholarly Writing Software- Cloud Based Using IBM Watson. User PortalThe User Portal made use of Microsoft Word? as the point of origin for all text analysis. Like other non-AI applications, such as Endnote, Grammarly, Refworks, the Scholarly Writing System incorporated an “add on” tool bar that may allow a user to initiate action, see Figure 4.Figure 4. Scholarly Writing Software-User Portal Toolbar. When the SWS tool buttons are activated (clicked) on the UP tool bar a “request” will be made to a program that is running in the background on the user’s computer. This will in turn call analysis functions into production via the Internet to the Watson Cloud. The background program will execute the required function on the user’s text and display the results in a second instantiation of Microsoft Word?. At this juncture the user will have the ability to immediately insert the altered text into the original text at the click of a button or reject the changes suggested by the SWS. A background program monitors the SWS actions that were taken for a particular document, and allow the user to access them in a Microsoft Word Windows?. UP functions will not be able to be accessed if the user is offline and no longer has a valid subscription to the SWS service. As a note, Microsoft Word for Mac? was not used in the prototype study. IBM Deep QAA prototype SWS was adapted using IBM Watson’s Deep QA architecture that was first introduced in the game show Jeopardy and uses a “massively parallel probabilistic evidence-based architecture” (Ferrucci, D.; et al. 2010 Deep QA, p. 68). Deep Q&A development requires a number of subject matter experts, IT systems engineers, and instructional designers to enable integration, function, assessment of the style rules as well as grammar, mechanics, and other content analytics. Figure 5 demonstrates how deep QA performs tasks. Figure 5. IBM Watson Deep QAIn Figure 5 the process starts out with a question analysis that is an attempt by a Watson based system to understand the question/query. Moving down the model query decomposition breaks the question down into smaller more easily answered by rule-based components. Hypothesis takes the results of question analysis and produces what is known as “candidate answers”. Soft filtering narrows down candidate answers. Synthesis ensures that each component question is created from the decomposition is completely answered. The process of final merging and ranking ensures that all answers derived from the original decomposition are recombined into a complete answer. In the final answer and confidence stage an answer is provided with a confidence level indication that can be in the form of a percentage. It should be noted that Watson’s Deep QA confidence level is rarely (if at all depending on the application) is at 100%. Confidence levels noted in the SWS prototype varied from a low in the 70th percentile to highs in the mid to upper 90th percentile. ConclusionArtificial augmentation intelligence/knowledge based systems are a reality. Major technology giants such as IBM, Google, Microsoft, Facebook, and others are expending significant amounts of time and money to further AAI. IBM is currently using the Watson AAI/Cognitive computing cloud-based platform in health care with significant success. The purpose of this qualitative single case study was to determine if what was learned in the health care profession using the IBM Watson architecture based on Watson Deep QA could be used to develop educational tools; in this case a prototype Scholarly Writing Software program. After interviewing 20 technology and educational professionals it was determined that it was feasible. A working prototype was then developed. It is the researchers hope that other researchers and developers will seek out new and innovate ways to utilize AAI in the education environment in the age of cognitive computing.Acknowledgements: The authors would like to thank Dr. Mansureh Kebritchi University Research Chair, for Educational and Instructional Technology Research (CEITR) at the University of Phoenix for awarding a research fellowship to conduct this study. We would also like to thank Sara Weber who thoughtfully edited and proofed the manuscript. Finally, to the 20 IT experts who gave their valuable time to provide their opinions and expertise. ReferencesAblin, J. A. (2008). Learning as problem design versus problem solving: Making the connection between cognitive neuroscience research and educational practice. Mind Brain and Education, 2(2), 52-54.Abrami, P. C., Bernard, R. M., Bures, E. M., Borokhovski, E., & Tamim, R. M. (2011). Interaction in distance education and online learning: Using evidence and theory to improve practice.?Journal of Computing in Higher Education,?23(2-3), 82-103. doi:, R., & Sajja, P. (2010). Knowledge-Based Systems. Sudbury, MA: Jones and Bartlett Publishers.American Psychological Association. (2010). Concise rules of APA style sixth edition. Washington, DC: American Psychological Association.Aytekin, I. I., AbdulAziz, A. d., Barakat, H. h., & Abdurrahman, A. d. (2012). Effectiveness of Instructional Design Model (Isman - 2011) In Developing The Planning Teaching Skills Of Teachers College Students' At King Saud University.?Turkish Online Journal Of Educational Technology,?11(1), 71-78.Banner, D. D., & Zimmer, L. L. (2012). Informed consent in research: An overview for nurses. Canadian Journal of Cardiovascular Nursing 22(1), 26-30. Retrieved from Barinka, A. (2013, ). IBM offers Watson as cloud tool. Journal - Gazette Retrieved from , R. M., Abrami, P. C., Borokhovski, E., Wade, A., Tamim, R., Surkes, M. A., et al. (2009). A meta-analysis of three interaction treatments in distance education. Review of Educational Research, 79(3), 1243–1289. doi:10.3102/0034654309333844v1.Coppola, N., W., Hiltz, S. R., & Rotter, N. (2002). Becoming a virtual professor: Pedagogical roles and ALN. Journal of Management Information Systems, 18(4), 169-170.Ferrucci, D. :. B.,E., Chu-Carroll, J., Fan, D., Gondek, D., Kalyanpur, A., Lally, A., . . . Welty, C. (2010) Building Watson: An overview of the Deep QA project. AI Magazine, 31 (3), 59-70. Retireved from . Ferrucci, D. :. B.,E., Chu-Carroll, J., Fan, D., Gondek, D., Kalyanpur, A., Lally, A., . . . Welty, C. (2010). The AI behind Watson — technical article Retrieved 3/23/2014, 2014, Retrieved from , T. Wilcox, R (2015). IBM Watson and Medical Records Text Analytics HIMSS Presentation [PowerPoint slides]. Retrieved from , P. (2013). Becoming an academic writer: 50 exercises for paced, productive, and powerful writing. Thousand Oaks, CA: Sage Publications.IBM Education (2016). Transform learning experiences?with Watson: Personalized learning on a whole new level. Somers, NY. IBM Corporation. Retrieved from Watson, IBM's jeopardy computer. (2011).?Information Week - Online,?Retrieved from , Y. (2009). Building a conceptual framework: Philosophy, definitions, and procedure. International Journal of Qualitative Methods, 8(4). Jones, R. (2014). Dissertation writing: The importance of alignment | the refractive thinker Retrieved from , J., & Kidd, T. (2010). Towards best practices in online learning and teaching in higher education. MERLOT Journal of Online Learning and Teaching, 6(2), 533-541Kellogg, R. T., & Raulerson, B. A.,III. (2007). Improving the writing skills of college students. Psychonomic Bulletin & Review (Pre-2011), 14(2), 237-42. Retrieved from III, J. (2015). IBM Computing, cognition and the future of knowing: How humans and machines are forging a new age of understanding [White paper]. Retrieved from , B., & Huberman, M. (1994). Qualitative data analysis: An expanded source book. Thousand Oaks, CA: Sage Publications.Minocha, S. (2009). Role of social software tools in education: a literature review. Education + Training, 51(5/6), 353-369. doi: doi:10.1108/00400910910987174.Morse, J.M. & Richards, L (2002). Readme first a user’s guide to qualitative methods. Thousand Oaks, CA: Sage Publicaitons.Oblinger, D. (2003). Boomers, Gen-Xers, Millennials: Understanding the new student. EDUCAUSE(July/August).Shead, S. (2016). Nick Bostrom: London’s DeepMind is winning the global race to develop human-level artificial intelligence. [Online exclusive]. Business Insider/Tech Insider. Retrieved from: , P. (2016, May 12). Announcing SyntaxNet: The world’s most accurate parser goes open source. Google Research Blog. Message posted to: , D. W., III (2010). Qualitative interview design: A practical guide for novice investigators. The Qualitative Report, 15(3), 754-760. Retrieved from , H., & Varol, C. (2014). Improving female student retention in computer science during the first programming course. International Journal of Information and Education Technology, 4(5), 394-398. doi:, K. (2013). IBM Watson: Revolutionizing healthcare??Young Scientists Journal,?6(13), 17-19. doi:'s Third New International Dictionary, Unabridged, s.v. “vaporware,” accessed April 18, 2015, , M. (2016). Why Facebook, Amazon, Microsoft, and Google all desperately need you to know that the robots are coming. [Online exclusive]. Business Insider/Enterprise. Retrieved from: , M. (2016). Microsoft releases MS MARCO dataset to train AI systems. [Online exclusive]. SiliconAngle. Retrieved from: , R., Crow, G., Heath, S., & Charles, V. (2008). The management of confidentiality and anonymity in social research. International Journal of Social Research Methodology, 11(5), 417-428. ................
................
In order to avoid copyright disputes, this page is only a partial summary.
To fulfill the demand for quickly locating and searching documents.
It is intelligent file search solution for home and business.
Related searches
- university of phoenix portal
- university of phoenix online sign in
- university of phoenix ecampus website
- university of phoenix ecampus portal
- university of phoenix majors list
- university of phoenix bachelor programs
- university of phoenix bachelors cost
- university of phoenix student portal
- university of phoenix ecampus log in
- university of phoenix site down
- university of phoenix biology degree
- university of phoenix employee portal