The intelDok project will bring natural language capabilities to our iberDok communications and document generation solution.
Currently, the modality for editing documents and forms based on templates, with the integration of static and dynamic objects, has proven to be one of the most outstanding on the market, but we wanted to go further.
The project consists of incorporating functionalities linked to natural language (LNP) to facilitate working with documents, from writing to migration, normalization or consumption. This goes through the development of the intelDok project, on which we are already working. With its implementation, the creation of documents, forms and templates will no longer be done with human interaction, but automatically.
intelDok will help to include the most convenient blocks of information based on the context, instead of relying on rules. In addition, it will generate predictive texts based on other templates and documents with which it shares tags, filling in complementary paragraphs, it will normalize the text for anomalous or infrequent words, and it will summarize and classify the documents. A search engine for pre-existing templates will also be developed using natural language that will map the documents by similarity.
According to our research, in the state of the art there are still no works that combine the technologies proposed by intelDok to create a platform that allows all these areas, which range from the generation of synthetic data sets, both from texts and images, to the automatic cataloging of documents and their summaries, through the automatic generation of templates based on the text and the context of the original textual information.
With the development of the intelDok project, we take the generation of documents to another level, incorporating innovative functionalities thanks to the possibilities offered by Artificial Intelligence and NLP natural language technologies, adding these innovations to others already implemented, such as Blockchain technology.