Turkish Named Entity Recognition
In information extraction, a named entity is a real-world object, such as a person, location, organization, product, etc., that can be denoted with a proper name.
Wikipedia
Named Entity Recognition
Named Entity Recognition (NER) is the task of determining the named entities in the given text. It is a non-trivial task since it is infeasible to cover all possible NEs in a simple lookup-list. Thus, Turkish.AI NER module depends on complicated Deep Learning models to identify and classify the types of entities in a sentence.
Turkish Named Entity Recognition Module
Built for developers in mind.
FINE-GRAINED TYPES
Do not limit yourself with just PERSON, ORGANIZATION and LOCATION entity types.
MORE THAN 30 ENTITY TYPES
ENTITY NORMALIZATION
Do not limit your applications with just named entity recognition.
EXTRACT THE ACTIONABLE STRUCTURED DATA
ENTITY LINKING
Coming soon.
ENTITIES MATCHED WITH KNOWLEDGE GRAPH ENTITIES
Fine-Grained Entity Types
Go beyond the usual person, organization and location entity types. More advanced NER systems like Turkish.AI have fine-grained classes such as product names, titles, event names and so on. Also, custom entity types are required for domain-specific solutions such as IBAN, Turkish Identity Number etc. which are also supported by our platform.
ORG
Organization names like companies, governmental and non-governmental organizations, sports clubs.
NORP
Nationalities, religious or political groups.
GPE
Geo-political entities e.g. everthing with a governing body such as countries, cities, villages and etc.
LOCATION
All geographical entities such as mountains, rivers, streets and other locational entities.
FACILITY
All human-made facilities like airports, bridges, buildings, roads, hospitals and similar.
PERSON
Each distinct person (real life or fictional) or set of people (such as family names).
TIMEX
Temporal expressions referring a specific time point, period or duration. May be absolute or relative where we need a reference date and time (such as document date time).
NUMBER
All types of numerals (cardinal, ordinal and distributional) expressed with digits, text or mixed expressions of digits and text. Normalization is also applied to get the integer of float value.
and
33
more
Entity Normalization
Detecting entities is good, normalizing them is even better.
Get the actionable values instead of raw text.
A few normalization examples...
See our Turkish Named Entity Recognition and Normalization module in action.
Monetary Expressions
Monetary values and their currencies are captured and normalized, even the ones specified as ranges.
Temporal Expressions
Both absolute and relative temporal expressions are normalized.
Law
Legislative entities such as laws and regulations are captured and normalized.
Measurement Normalization
Measurements like lenght, weight and many more are recognized and normalized.
Do you want to try your own examples ?
Check our demo page to see Turkish.AI in action.