SlideShare a Scribd company logo
1 of 4
1Confidential
NER ?
Name Entity ,Relation and Events
Named Entities:

Person

Organization

Designation

Location

Date Time

Money
Salary
Net profit
Hike
Gross profit
Turnover

Percent
Hike rate
Growth rate
Increase by
Relations :

Person

→ Organization

→ Designation

→ Money

→ Percent

→ Quoted Text

→ Location

Organization
→ Percent

→ Quoted Text

→ Money

→ Location
Events :

5WH

Who → Person

What → Event description

Whom → Organization

When → Event date time

Where → Location
Information :

Person :->{title}->Organization

Person :->{title}->Designation

Person :->{title}->Money

Person :->{title}->Location

Organization :->{title}->Location

Organization :->{title}-> Percent

Organization :->{title}-> Percent
NER Consist :

Gazetteer Lookup

JAPE Grammers

Co - reference
2Confidential
Example
TCSTCS
CEO
N Chandrasekaran
Person :->{title}->Organization
WiproWipro
CEO
TK Kurian
TCSTCS
Managing Director
N Chandrasekaran
InfosysInfosys
CEO
Vishal Sikka
Person title Organization
N Chandrasekaran CEO TCS
Vishal Sikka CEO Infosys
TK Kurian CEO Wipro
3Confidential
Knowledge Network
Organiz
ation
Pe
rso
n
Desig
nation
Loc
atio
n
Mo
ne
y
Per
cen
t
TCS
Financial
News
TC
S
Tata Consultancy
Service
15 lakh
10 lakh
Rs 3,596.9 core
50 percent
21 percent
Kolkata
CEO
child
child
HikeBy
child
child
child
child
child
child
Growth rate
Salary
Net profit
N
Chandrasekar
an
Chandrasekar
an
child
child
hold_position
same_as
known_as
located_in
child
hike_rate
net_profit
4
ExtractionPipelineReasoner Application
Relation
Model
Inf Node
Model
Rule Set 2Rule Set 1
Inf Relation
Model
Rule Set n
Inf
Document
Model
Tokenizer
Sentence
Spiltter
POS tagger
Customize
Gazetter
Co-refencer
How ??
Finance Gazetteer
Relation
Transducer
Document
Model
Node Model ….........................................
Knowledge
Store
KnowledgeAccessAPI
Application
Relation schema
KnowledgeAccessSPARQLAPI
Asserted
Network
Inferred
Network

More Related Content

Viewers also liked

1. Development Pro Forma Finished
1. Development Pro Forma Finished1. Development Pro Forma Finished
1. Development Pro Forma FinishedRebecca Coughlin
 
Stakeholder Apoio em Comunicação
Stakeholder Apoio em ComunicaçãoStakeholder Apoio em Comunicação
Stakeholder Apoio em ComunicaçãoIvan Margarido
 
гибкая методология разработки по
гибкая методология разработки погибкая методология разработки по
гибкая методология разработки поpoverhnost
 
Pesos e notas míninas em todos os cursos
Pesos e notas míninas em todos os cursosPesos e notas míninas em todos os cursos
Pesos e notas míninas em todos os cursosJornal do Commercio
 

Viewers also liked (11)

1. Development Pro Forma Finished
1. Development Pro Forma Finished1. Development Pro Forma Finished
1. Development Pro Forma Finished
 
Net framework
Net frameworkNet framework
Net framework
 
Slideshare
SlideshareSlideshare
Slideshare
 
Stakeholder Apoio em Comunicação
Stakeholder Apoio em ComunicaçãoStakeholder Apoio em Comunicação
Stakeholder Apoio em Comunicação
 
jejemon.
jejemon.jejemon.
jejemon.
 
гибкая методология разработки по
гибкая методология разработки погибкая методология разработки по
гибкая методология разработки по
 
Navarra
NavarraNavarra
Navarra
 
Pesos e notas míninas em todos os cursos
Pesos e notas míninas em todos os cursosPesos e notas míninas em todos os cursos
Pesos e notas míninas em todos os cursos
 
Veganuary presentation
Veganuary presentationVeganuary presentation
Veganuary presentation
 
Task 1 Page Layout
Task 1 Page Layout Task 1 Page Layout
Task 1 Page Layout
 
Somália
Somália Somália
Somália
 

Contextual Entity detection and semantic text-KSS

  • 1. 1Confidential NER ? Name Entity ,Relation and Events Named Entities:  Person  Organization  Designation  Location  Date Time  Money Salary Net profit Hike Gross profit Turnover  Percent Hike rate Growth rate Increase by Relations :  Person  → Organization  → Designation  → Money  → Percent  → Quoted Text  → Location  Organization → Percent  → Quoted Text  → Money  → Location Events :  5WH  Who → Person  What → Event description  Whom → Organization  When → Event date time  Where → Location Information :  Person :->{title}->Organization  Person :->{title}->Designation  Person :->{title}->Money  Person :->{title}->Location  Organization :->{title}->Location  Organization :->{title}-> Percent  Organization :->{title}-> Percent NER Consist :  Gazetteer Lookup  JAPE Grammers  Co - reference
  • 2. 2Confidential Example TCSTCS CEO N Chandrasekaran Person :->{title}->Organization WiproWipro CEO TK Kurian TCSTCS Managing Director N Chandrasekaran InfosysInfosys CEO Vishal Sikka Person title Organization N Chandrasekaran CEO TCS Vishal Sikka CEO Infosys TK Kurian CEO Wipro
  • 3. 3Confidential Knowledge Network Organiz ation Pe rso n Desig nation Loc atio n Mo ne y Per cen t TCS Financial News TC S Tata Consultancy Service 15 lakh 10 lakh Rs 3,596.9 core 50 percent 21 percent Kolkata CEO child child HikeBy child child child child child child Growth rate Salary Net profit N Chandrasekar an Chandrasekar an child child hold_position same_as known_as located_in child hike_rate net_profit
  • 4. 4 ExtractionPipelineReasoner Application Relation Model Inf Node Model Rule Set 2Rule Set 1 Inf Relation Model Rule Set n Inf Document Model Tokenizer Sentence Spiltter POS tagger Customize Gazetter Co-refencer How ?? Finance Gazetteer Relation Transducer Document Model Node Model …......................................... Knowledge Store KnowledgeAccessAPI Application Relation schema KnowledgeAccessSPARQLAPI Asserted Network Inferred Network