The document summarizes the topics discussed at an XAPI Chinese CoP meeting in February 2016. It covered the XAPI vocabulary specification, linked data/semantic web, linked data in education and content recommendation, semantic search and Google Knowledge Graph, monetizing data and adding intelligence. It also included a case study on Hong Ding Educational Technology using XAPI data and partnerships to provide differentiated learning paths. The document emphasized collaborating on standards for competency, user data, content metadata and xAPI statements to enable partnerships and monetizing data while ensuring security, regulation and collective decision making.
2. Topics
● XAPI Vocabulary spec. From ADL
● Linked Data / Semantic web. / Web 3.0
● Linked Data in education and content recommender
● Semantic search and Google Knowledge Graph
● APIs eat software (connect with partners and services)
● How should we exploit data and build intelligence layer?
● Case Study (Hong Ding Educational Technology)
● Monetize your data and add value (intelligence)
3. Draft v1.0 published in Feb. 2016:
Vocab Spec. - gitbook: http://bit.ly/read-vocab-spec
Vocab Primer - gitbook: http://bit.ly/read-vocab-primer
“In addition to structural interoperability, semantic data interoperability is also
needed for both humans and applications to meaningfully interpret the information
being exchanged. “
AcrossX vocab registry is the first vocabulary registry besides ADL & IEEE ADB
registries that completely follows the xAPI Vocab. Spec. : (by Jessie Chuang, Jason
Haag) : https://w3id.org/xapi/acrossx
ADL xAPI Vocabulary Spec.
4. “MUST” in xAPI Vocabulary Spec.
Anyone establishing new vocabularies and new vocabulary terms for xAPI MUST use
HTTP IRIs so that they can be resolved/dereferenced.
Anyone creating new vocabulary terms MUST provide descriptive metadata by
minimally using the recommended classes and properties in this specification, and
associate the terms with a vocabulary (concept scheme).
Anyone establishing a new vocabulary MUST publish a human-readable
representation of the vocabulary dataset as HTML accessible at the IRI.
Anyone establishing a new vocabulary MUST publish at least one additional machine-
readable representations of the vocabulary dataset in RDF (e.g., RDFa, JSON-LD,
Turtle).
5. Why Linked Data?
The ability to link data from diverse sources is a motivator for many projects, as
different CoPs seek to take advantage of semantically rich data that was previously
spread across disparate sources. By adopting W3C’s RDF standard as the data model
for xAPI vocabulary resources and their metadata, the xAPI specification can
potentially gain a whole new level of precision for machine readability and semantic
interoperability. Upon implementation of this vocabulary specification (and
refinement of vocabulary publishing practices) the xAPI community will benefit from
new capabilities - open up new doors for improved learning analytics, federated search
(of vocabulary), dynamic look-up of xAPI vocabulary data within authoring
applications, improved discovery and reuse of xAPI vocabularies, multilingual
translation, and basic natural language processing capabilities. -- Jason Haag, ADL
6. Linked Data / RDF - Resource Description Framework
RDF and JSON-LD work in a simple way of dividing information into triples:
“subject – predicate – object.”
RDF / JSON-LD 是 W3C 的语义网 (Semantic Web) 资料模型标準,其 Graph Data
Model 是 NoSQL 的一种,基本组成成分为 Subject – Predicate – Object,称为
triples,叁者都是以 URI 来代表,Subject 与 Object 是任何可被识别的实体或资讯类
对象 (nodes),Predicate 代表这两者的关係 (edge, link)。
这样极简的构造允许结构或非结构化的资料数据互相整合,不须事先规定相同架构,
特别适合分岐、分散不同来源的大数据(Big Data)。
Linked Data 是语义网 (Semantic Web) 的正确实践。主要目的是让机器可读懂
(machine-interpretable) 所传递的资料,以发展人工智慧。
7. A cornerstone of the semantic web is its use of newer graph-based approaches
and technologies – such as the RDF and SPARQL W3C initiatives.
8. For most types of data storage, there is the concept of some elements of data (whether they be for
example data nodes or data tables) having more precedence, or importance, over other elements.
(image from http://www.linkeddatatools.com/semantic-web-basics)
9. ● A shift in thinking from publishing data in human readable HTML
documents to machine readable documents.
● It provides a way to homogeneously integrate heterogeneous resources,
taking all that information published in HTML documents in different places,
and allow it all to be treated - and researched - as if it were one database.
● Opening up the web of data to artificial intelligence processes (getting the
web to do thinking for us).
● Encouraging businesses to use data already available on the web (e.g. DBpedia
from all Wikipedia data, read Taking Advantage of Existing Linked Data )
Semantic Web : Web 3.0
14. Semantic Computing (SC)
It’s to represent concepts and their relations in an integrated semantic network
that loosely mimics the inter-relation of concepts in the human minds. The
knowledge is represented by ontologies and can be used to annotate data and
infer new knowledge.
SC plays a crucial role in dealing with multi-sensory and multi-model
observations, leading to integration of data from diverse sources.
SC has a rich history of 15 years, resulting in various annotation standards for a
variety of data. The annotated data is used for interpretation by cognitive &
perceptual computing.
15. Semantic Search
Semantic search is to improve search results by focusing on user intent and
how the search relates to other information in a wider sense, or its contextual
relevance -- what a user really means.
The Knowledge Graph (launched May, 2012) is Google’s database of
information. The information is gained by crawling the web. Google indexes
and organizes this information (known as graphing), rendering it useful to
queries and commands. Google becomes more of a "knowledge engine" rather
than the traditional "information engine". (Forbe, Nov. 2015)
Hummingbird (launched Sep., 2013) was essentially an entirely revamped
version of Google’s search algorithm.
BTW, Customizing Your (company’s) Knowledge Graph
19. Google Knowledge Graph API (to replace Freebase API)
The Knowledge Graph Search API (launched Dec. 2015) lets you find entities in the
Google Knowledge Graph. The API uses standard schema.org types and is compliant
with the JSON-LD specification.
Some examples of how you can use the Knowledge Graph Search API include getting
a ranked list of the most notable entities that match certain criteria, annotating /
organizing content using the Knowledge Graph entities.
The Knowledge Graph has millions of entries that describe real-world entities like
people, places, and things. These entities form the nodes of the graph. The following
are some of the types of entities found in the Knowledge Graph: Book, BookSeries,
EducationalOrganization, Event, GovernmentOrganization, LocalBusiness …. etc.
20. Learner
Learning DesignerData Analyst
● Interoperability
● Analytics (algorithms,
machine learning API)
● Recommender (context
matching, adaptive/ZPD)
● Learning theories
● Psychology
● Design (experience design, UI)
● Subject Matter Expertise
*Engagement *Visualization as cognitive agent
*Navigator (recommender)
*Context !!!
inform iterate
co-design
integrate
*Community & social cues
*Personalized assistance
Gamified layer
on data from
across platforms
Big Data
[Data-Driven Learning]
22. Methodology of
Leveraging Big Data
It’s been used in...
Lean Startup Prototyping
Marketing Analytics
Business Intelligence
Rapid Cycle Tech Evaluations(DOE)
Data-Driven Learning Design
Data is communication !
(with human’s intelligence)
25. Learning
Planning
Activity/Agent Profile API
IRS, Quiz service
Integration of Workflow and Data flow *Adaptive design
*Branches/options
Videos or MOOC platform
Mobile Apps
IoT
Wearable
AR
VR ...
Real behavior data sent to
Game/ Gamification/ 3rd
platforms for rewards
26. Amazon --- “Anyone who doesn’t do this will be fired. “
Amazon.com is one of the best examples of a programmable enterprise. The company evolved
from an online bookseller to a leader in cloud computing and infrastructure-as-a-service in
large part because of a mandate set forth by its CEO Jeff Bezos in 2002.
He proclaimed that all services within the company must be built in a way that they could be
exposed to one another to enable simple communications. In addition, all of these so-called
service interfaces must be planned and designed to enable exposure to outside developers.
Bezos couldn’t foresee which of these services would end up being externalized. But the move
transformed Amazon internally and led to some products that went far beyond online
bookselling: Amazon’s Elastic Compute Cloud and its relational database services, among
others. Read about the story rant by an ex-Amazon-now-Google-employee.
Software is eating the world (Marc Andreessen, 2011), APIs eat software.
27. Clients can listen for statements being pushed to them from the LRS.
Webhooks are a way to register client applications that are interested in certain sets
of statements. Every webhook that is registered contains the client endpoint as well
as filters that dictate what kind of statements the client wants the LRS to send.
A WebHook is an HTTP callback: a simple event-notification via HTTP POST. A
web application implementing WebHooks will POST a message to a URL endpoint
registered when certain things happen. (www.webhooks.org)
Webhooks
28.
29. Images from Hong Ding Edu. Tech.
Embedded VisCa
Dashboard for
teachers or learners
31. XAPI data
Whenever K fails practice
A in Hong Ding, VisCa
will notify 1Know
Assessment and practice service
Light Learning Management System:
1Know
1Know pushes a helpful
video or resource link to K
This is K
XAPI data
Whenever K
finishes watching
video B, VisCa will
notify Hong Ding
Hong Ding pushes an
appropriate practice
set to K
Feedback
Request
Interact ..
Seamless Partnership
Automate differentiated learning paths, which can be designed by instructors or adapted by algorithms.
You know better about your learners !
32.
33. Monetize Your Data and Add Values (intelligence) !
● Let’s collaborate for consistency in:
○ Competency standards
○ User management data models
○ Content management model & metadata
○ Behavior events - xAPI statements recipes
● API call charge model (query, statistics, document of learner profile…)
● Security
○ Technology
○ Regulation and management
● By collective decisions of CoP participants