It’s a solution based on an algorithm that analyses text content and understands what are the relations in that content and how they link.
Our customer base range from Governmental institutions to Online Media Publications
There are several advantages for implementing a solution: -A fully Automated link creation system that doesn't depend on human curation selects the best related content (or the latest, or any other parameter) and frees up time for content producers to do what they do best. -Links and connections are always up-to-date to the latest documents created, linking automatically for any content created in the future. (no need to review past content to update links manually) -Connections sorted by relatedness, ensuring that the best content will get the most attention.
For document management applications, it can be very useful where there are a lot of technical jargon to be able to quickly identify what other documents are mostly related with each one of the terms, or if there are very similar terms.
Online News and Media CompaniesMedia Companies are great at generating content, but often don't have the resources to interlink the vast amounts of articles they produce. Generating links automatically between them solves many of these issues. We have identified 5 problems we can solve for them, click here to know more
Governments and ONG'sCreating a tight network of interconnected concepts and documents is fundamental to understand the context of laws, politicians and concepts.
Legal documents / patentsDiscovering related legislation or patents is a time consuming task. With Luma7 we automate this process and help the discovery process across millions of documents instantly.
Entity RecognitionLuma7 is able to identify and mark in the content more than 20 million Entities (concepts composed by multiple words) and classify them.
Text ClassificationOur algorithms are able to classify text and to organize by categories. By doing this we are able to identify what's relevant and how to treat that information.
Multiple LanguagesLuma7 can work with content in multiple languages in the same page, and process them independently to give results that are language relevant.
Model ParametrizationHumans write content in many different ways. In order to have better results we first clean and parametrize all text to transform it in a model as precise as possible.
Human InterfaceEach application has different requirements,
Frequent Data UpdateWe tried to make very high-quality product and so our code is very neat and clean. Whatever anyone could improve and modify the template to your liking.
The algorithm then generates a model space where it can calculate how closely related each word is arithmetically
From that we can estimate not only concept-to-concept relations, but also concept-to-article, paragraph-to-article and article-to-article.
Automatically Generated Link ConnectionsNo extra work for content producers to select and create html links
Ranking Links by "Distance"Our algorithms can understand what are the mostly related links to display by conceptual distance / frequency / date and other parameters.
Links permanently updatedBy continuosly analysing all web content, we are able to keep every link relevant and up-to-date