Text Mining and Knowledge Graphs in the Cloud: the Self-Service Semantic Suite (S4)

Text Mining and Knowledge Graphs in
the Cloud: The Self-Service Semantic
Suite (S4)
A webinar with
Marin Dimitrov, CTO of Ontotext
Feb 26th, 2015
Text Mining & Knowledge Graphs in the Cloud with S4 #1Feb 2015

• Semantic technologies for data management
• Self-Service Semantic Suite (S4)
• Text analytics
• RDF data management in the Cloud
• Knowledge graphs
• S4 for developers
• Roadmap
• Q&A session
Today’s Topics

About Ontotext
• Provides products & solutions for content
enrichment and metadata management
– 70 employees, head quartered in Sofia (Bulgaria)
– Sales presence in London, Washington & Boston
• Major clients and industries
– Media & Publishing
– Health Care & Life Sciences
– Cultural Heritage & Digital Libraries
– Government
– Education

Some of our clients

Semantic Technologies for
Smart Data Management

• How can we unlock more insight from text?
• How can we interlink & search across text and
structured data sources?
• How can we improve data & content reuse?
• How can we integrate data sources faster?
• How can we reuse external open data sources?
• How can we discover relations between entities?
Typical challenges for our customers

Ontotext’s vision for smart data
management
Graph Database
• Flexible RDF graph
data model
• Ontology metadata
layer
Semantic Search
• Semantic,
exploratory search
• Metadata driven
content
Text Mining & Interlinking
• People, locations,
organisations, topics
• Discover implicit
relations
• Reuse open knowledge
graphs

Ontotext and AstraZeneca
Profile
• Global, Bio-pharma company
• $28 billion in sales in 2012
• $4 billion in R&D across three continents
Goals
• Efficient design of new clinical studies
• Quick access to all of the data
• Improved evidence based decision-making
• Strengthen the knowledge feedback loop
• Enable predictive science
Challenges
• Over 7,000 studies and 23,000 documents
are difficult to obtain
• Searches returning 1,000 – 10,000 results
• Document repositories not designed for
reuse
• Tedious process to arrive at evidence
based decisions
#8Text Mining & Knowledge Graphs in the Cloud with S4 Feb 2015

Ontotext and LMI
Profile
• Established in 1961 to enable federal
agencies
• Specializes in logistics, financial,
infrastructure & information management
Goals
• Unlock large collections of complex
documents
• Improve analyst productivity
• Create an application they can sell to US
Federal agencies
Challenges
• Analysts taking hours to find, download
and search documents, using inaccurate
keyword searches
• Needed a knowledge base to search
quickly and guide the analysts – highly
relevant searches
• Extracts knowledge from collection of
documents
• Uses GraphDB to intuitively search and filter
• More than 90% savings in analyst time
• Accurate results

Ontotext and Euromoney
Profile
• Euromoney Institutional Investor PLC, the
international online information and events
group
Goals
• Create a horizontal platform to serve 100
different publications / 80 business units
• create a new unified publishing and
information platform
Challenges
• Different domains covered
• Sophisticated content analytics incl.
relation, template and scenario extraction
• Text analytics of reports and news in various
domains
• Extraction of sophisticated macro economic
views on markets and market conditions
• Triplestore for flexible data integration &
reasoning
• Multi-faceted search

The Self-Service Semantic Suite
(S4)

• Unlock the value of semantic technologies to SMEs
– Most success stories so far come from bigger companies
• Lower the technology adoption barriers and risks
– Challenge: perceived risks associated with new
technology adoption
– Challenge: insufficient resources to implement new
technologies
– Challenge: bureaucratic budgeting, procurement &
provisioning processes
Why did we create S4?

• Self-service capabilities for text analytics, content
enrichment and metadata management
– Text analytics for news, life sciences and social media
– RDF graph database as-a-service
– Access to large open knowledge graphs
• Available anytime, anywhere
– Simple RESTful services
• Simple, pay-per-use pricing
– No upfront commitments
What is S4?

• Utilise semantic technology for smart data
applications
– Extract more value hidden in text
– Interlink structured and unstructured data sources
– Semantic search (instead of keyword-based search)
– Reuse open knowledge graphs
• Low adoption cost and risk
• No need for complex planning & procurement
• Pay only for what you use, reduce TCO
S4 benefits

• Enables quick prototyping & shorter time-to-
market, increase innovation speed
• Available on-demand in the cloud, no provisioning
& operations required
• Based on enterprise grade semantic technology by
Ontotext
• Migration path from S4 based prototypes to
customised enterprise solutions with Ontotext
technology
S4 benefits

• Instantly available
• Free tier
• Easy to start, shorter learning curve
– Various add-ons, SDKs and demo code
• Simplify the technology stack for smart data
applications
• Focus on building applications, don’t worry about
infrastructure & operations
• Quicker prototyping, shorter development cycles
S4 for developers

Text Analytics

• Text analytics services
– News annotation
– News categorisation
– Biomedical
– Twitter
• Entity linking & disambiguation
– Mappings to DBpedia & GeoNames instances
– Mappings to biomedical data sources (LinkedLifeData)
• HTML, MS Word, XML, plain text input
• Simple JSON output
Text analytics with S4

• Entity types
– Person
– Organization
– Location
– Relation (affiliation, customer, competitor, partner,
acquisition, role, …)
– Keywords and key phrases
• Enterprise grade technology
– Based on successful text mining solutions for big media
& publishing companies
News analytics with S4

News analytics with S4

News analytics example
S4 result

News analytics example
API_KEY=s4trm64sb76u
KEY_SECRET=lrcki2kkajslsp6
SERVICE_ENDPOINT="https://$API_KEY:$KEY_SECRET@text.s4.ontotext.com/v1/news"
CONTENT="President Barack Obama is urging parents to get their children vaccinated in the face
of a measles outbreak that has infected more than 100 people in the United States. In excerpts
from an interview with NBC News that will air on Monday, Obama said measles was a
preventable disease."
CONTENT_TYPE="text/plain"
JSON_REQUEST="{"document" : "$CONTENT", "documentType" : "$CONTENT_TYPE"}"
curl -X POST -H "Content-Type: application/json" -d "$JSON_REQUEST" $SERVICE_ENDPOINT
{
"document" : "President Barack Obama is urging parents to get their children vaccinated in
the face of a measles outbreak that has infected more than 100 people in the United States.
In excerpts from an interview with NBC News that will air on Monday, Obama said measles
was a preventable disease" ,
"documentType" : "text/plain"
}
API key pair REST service
text
Request structure
Request
structure

• 17 top-level categories from the IPTC Subject
Reference System
– Arts / Culture / Entertainment, Crime / Law / Justice,
Disaster / Accident, Economy / Business / Finance,
Education, Environment, Health, Politics, …
– Based on successful text mining solutions for big media
& publishing companies
News classification with S4

News classification example
S4 result

News classification example
API_KEY=s4trm64sb76u
KEY_SECRET=lrcki2kkajslsp6
SERVICE_ENDPOINT="https://$API_KEY:$KEY_SECRET@text.s4.ontotext.com/v1/news-classifier"
CONTENT_URL="http://www.theguardian.com/world/2015/feb/04/taiwan-plane-crash-lands-in-
river"
CONTENT_TYPE="text/plain"
JSON_REQUEST="{"documentUrl" : "$CONTENT_URL", "documentType" :
"$CONTENT_TYPE"}"
curl -X POST -H "Content-Type: application/json" -d "$JSON_REQUEST" $SERVICE_ENDPOINT
{
"documentUrl" : "http://www.theguardian.com/world/2015/feb/04/taiwan-plane-crash-lands-
in-river" ,
"documentType" : "text/html"
}
API key pair REST service
URL Request structure
Request
structure

• 130 biomedical entity types
– Organism, Virus, Animal, Anatomical Structure, Organ,
Tissue, Cell, Genome, Chemical, Lab Result, Clinical Drug,
Biologic Function, Organ Function, Disease/Syndrome, …
– Based on successful text mining solutions for big
pharmaceuticals and healthcare providers
Biomedical analytics with S4

Biomedical analytics with S4

Biomedical analytics example
S4 result

• Entity types
– Person, Location, Organisation, Date, Address, Money
– Hashtag, Emoticon, URL, @UserID
• Based on TwitIE microblog pipeline by GATE /
University of Sheffield
Twitter analytics with S4

Twitter analytics example

RDF Data Management

• Standards compliance
– Based on a mature set of W3C standards: RDF/S, OWL,
SPARQL
– Portability & interoperability
• Schema-less data integration, easy querying of
diverse data
• Complex & exploratory queries
• Infer implicit relations in the graph
• Reuse open knowledge graphs (Linked Open Data)
RDF for smart data management

A visual view of RDF data
Sub-properties
Sub-classes
Transitive
relations
Inference

• High performance RDF database
• Full SPARQL 1.1 support
• Various reasoning profiles, including custom rules
• Efficient data integration (“sameAs” optimisations)
• Efficient deletion of statements & their inferences
• Geo-spatial indexing & querying with SPARQL
• RDF Rank, full-text search, 3rd party plugins
GraphDB by Ontotext

• Ideal for customers who are…
– still evaluating and testing RDF technology
– In the early phase of adoption / POC
• Enterprise grade RDF database in the Cloud
– No need for upfront payments for licenses & hardware
– Pay only for what you use, when you use it
– Instantly operational within minutes
– No need for complex planning - use as many DB
instances for as long as needed
– Timely upgrades to the latest version
• Self-managed and full-managed options
RDF database in the Cloud with S4

• Available from AWS Marketplace
• Variety of hardware configurations
– 2 to 8 CPU cores / 8 to 61 GB RAM
– IOPS performance & encryption (EBS)
• Manage large data volumes
• Pay-per-hour pricing
Self-managed database in the Cloud

• (available in Q2’2015)
• Low-cost DBaaS available 24/7
• Ideal for small & moderate data volumes
• Instantly start new databases when needed
• Zero administration: automated operations,
maintenance & upgrades
• Users pay only for the actual database utilisation
– database size + number of queries per period
Fully-managed database in the Cloud

Knowledge Graphs

• SPARQL query endpoint to FactForge knowledge
graph
– 500 million entities
– 5 billion triples
• Key LOD datasets integrated
– DBpedia, Freebase, GeoNames, WordNet
– Dublin Core, SKOS, PROTON ontologies and
vocabularies
Knowledge graphs with S4

Knowledge graph query example
SPARQL query
using DBpedia
data

For Developers

Getting started in minutes
1. Register a personal
account at s4.ontotext.com
2. Generate an
API key pair
3. Check out the docs,
demos & code at
docs.s4.ontotext.com
4. Contact us
with questions!

• Java & C# SDKs
• Sample code
– Java, C#, NodeJS, JavaScript, Python, PHP, Groovy
– Curl examples for the most impatient
• GATE plugin (UIMA plugin in Q2’2015)
• Firefox plugin
• Online documentation
S4 for developers

• March 1st – 30th 2015
• Submit a cool text analytics & Linked Data
application using S4
• $1,000 for the winning submission
• More details at http://bit.ly/s4-challenge
S4 Developers Challenge

Roadmap

• Text analytics
– Multi-lingual text analytics
– Sentiment analytics
– JSON-LD output format
• RDF databases
– Fully managed RDF DBaaS
– Regular updates of the self-managed GraphDB on AWS
• Knowledge Graphs
– Private knowledge graph databases with
DBpedia/Wikidata
– 3rd party Linked Data visualisation & exploration tools
What to expect in 2015?

• Pricing plans
– Simple, transparent, usage based pricing
– Pay only for what you use, when you use it
• For developers
– UIMA plugin for S4
– More SDKs
– mode add-ons
– Demos and sample code
– S4 Developers Challenges
What to expect in 2015?

Key Takeaways

• Semantic technologies provide good capabilities
for smart data management
• Key S4 benefits
– Lowers the risks and costs for semantic technology
adoption
– Shortens time-to-market, reduces TCO
– Provides a safe migration path into custom enterprise
solutions with Ontotext technology
• Key S4 capabilities
– Various text analytics components (more to come!)
– Self-managed & fully managed RDF DB in the Cloud
– Knowledge graphs with reusable open data
Key Takeaways

• Online documentation
– http://docs.s4.ontotext.com/
• Sample code & demos on GitHub
– https://github.com/Ontotext-AD/S4
• Helpdesk
– http://support.s4.ontotext.com/
• Twitter
– @Ontotext_S4
Additional S4 resources

Thank you!
Text Mining and Knowledge Graphs in the Cloud:
The Self-Service Semantic Suite
A link to the recording will be sent out shortly
Feb 26th, 2015

Text Mining and Knowledge Graphs in the Cloud: the Self-Service Semantic Suite (S4)

Recommandé

Recommandé

Contenu connexe

En vedette

En vedette (11)

Plus de Marin Dimitrov

Plus de Marin Dimitrov (14)

Dernier

Dernier (20)

Text Mining and Knowledge Graphs in the Cloud: the Self-Service Semantic Suite (S4)