ENTITY LINKING
AT SPEED AND AT SCALE
If you are visiting this site, you are probably already familiar with Natural Language Processing, often referred to as NLP.

If that’s not the case, click here to see how Wikipedia defines NLP.
Natural language processing (NLP) is a subfield of linguistics, computer science, and artificial intelligence concerned with the interactions between computers and human language, in particular how to program computers to process and analyze large amounts of natural language data. The result is a computer capable of "understanding" the contents of documents, including the contextual nuances of the language within them. The technology can then accurately extract information and insights contained in the documents as well as categorize and organize the documents themselves.
NLP is in fact an umbrella term for several computing and analysis tasks.

Chief amongst them is ENTITY LINKING, as it helps connects words and sentences to their actual meaning. At the risk of using technical jargon, most NLP tasks relate to the lexical and syntactic dimensions of language - while Entity Linking is about semantics.

Unlike many NLP toolkits and commercial APIs who do not shine with Entity Linking, ENTITYZE focuses on this one task only.

Not only does ENTITYZE offer better and faster Entity Linking, it comes with a few extras that you can't find anywhere else.

DIFFERENTIATORS
(FEATURES)

MORE INPUT FORMATS

With more than a thousand file formats supported, ENTITYZE can process text coming from almost any data source. With other tools, you have to do this conversion yourself.

MORE OUTPUT OPTIONS

Unlike other tools that provide one-size-fits-all results, ENTITYZE lets you customize the content and granularity of outputs to your exact needs.

MORE CONNECTORS

ENTITYZE comes with connectors that bring a seamless integration with your existing data pipeline tools and that let you leverage your existing ontologies or information graphs.

MORE VISUALIZATIONS

ENTITYZE DOC2GRAPH is a standard feature that lets you visualize the connections inside a collection of text documents. You can add or remove detail with a simple click.
DIFFERENTIATORS
(TECHNOLOGY)

GRAPH SIZE AND DIVERSITY

We combine and organize up-to-date knowledge graph data from several public and semi-public sources, resulting in tens of millions of interconnected entities.

MORE PRECISE THAN VECTORS

While word vectors and transformers have arguably changed the face of NLP, they do not deliver state-of-the-art results for the specific task of entity linking.

PRE-COMPUTED DATA

For maximum speed, we pre-compute, optimize and compress trillions of data points amounting to TBs of data and we then fit key results in memory for ultra-fast response.

PROPRIETARY DISAMBIGUATION

There is a usual trade-off between speed and accuracy when it comes to entity linking. Our #1 R&D focus was to develop a disambiguation algorithm without any compromise.
FREQUENTLY ASKED
QUESTIONS

HOW MANY LANGUAGES?

So far, six overall. ENTITYZE supports documents in English with a slightly more advanced level of precision than documents in these five other languages: Spanish, Portuguese, French, Italian and German.

WHAT ABOUT ACCESS RIGHTS?

Each user or team receives a unique access token. You can monitor and restrict usage in real-time via a standard API gateway dashboard.

WHAT ABOUT CONFIDENTIALITY?

When ENTITYZE resides in your private cloud environment (recommended deployment scenario), none of your proprietary data or content is ever sent out to our servers. If a troubleshoot need arises, you are in full control of what debugging data you send us and what debugging data you choose to keep.

WHAT ABOUT DATA SECURITY?

The ideal deployment scenario for ENTITYZE is a sandboxed, containerized application within your private cloud environment. All your existing security standards and protocols are applied to ENTITYZE. From time to time, you have the option to install some data and code updates - it is a process that you initiate as we do not have remote access to your environment after installation.

WHAT ABOUT ONTOLOGIES?

You can connect your own company-specific or business domain ontologies to the existing ENTITYZE knowledge graph. In fact, more than 500 distinct ontologies can co-exist inside each installation of ENTITYZE. Why so many? It allows us to split an ontology in multiple parts if needs be - this is helpful when you need to apply very granular prioritization rules to the entity linking process.

WHAT ABOUT SENTIMENT ANALYSIS?

While ENTITYZE can identify and detect words that express sentiments, the system cannot reliably detect instances of sarcasm at this time. If internal results improve to the point where we can deliver more dependable results than the market, we will add sentiment analysis to our service offering.
TRY ENTITYZE API FOR FREE