Revolutionary AI

Challenge what’s possible with our unique layout-aware Language Model (LAMBERT) and 2D Contextual Awareness technology

2D Contextual Awareness

Applica’s 2D Contextual Awareness leverages machine learning in the form of deep neural networks when applying algorithms from computational linguistics and computer vision. This unique technology mimics the way a human works with documents containing various layouts (e.g.: forms, tables, reports, etc.) and considers both textual and graphical aspects before finalizing results. Coverage of both channels of information enables precise semantic analysis and information extraction, which in turn reduces the workload for some business processes by 90% and decreases the error rate by 85%.

Layout-aware Language Model (LAMBERT)

The most effective NLP methods are based on neural language modeling, however all the language models that are currently used are one-dimensional, and they are blind to text structure. As a result, Applica’s layout-aware models beat all 1D models in challenges that deal with real business documents: a mix of semi-structured and unstructured text. For instance, when interpreting a table, it’s necessary to take into account its structure and the correct interpretation of a table involves understanding what is a column, row, cell, etc.

Technology benefits

Most AI requires supervised machine learning using large volumes of annotated documents. Thanks to our progressive neural language modeling techniques, Applica reduces supervised learning to a minimum.


SemEval 2020 Task 11

We are pleased to announce that Applica sits at the #1 and #2 spots on the leaderboard for SemEval’s 2020 task 11, an objective industry test aimed at detection of propaganda techniques in news articles. The team that developed Applica’s 2D Contextual Awareness applied our proprietary technology in a similar manner for this project. Wins such as this, against some of the best minds in the world, highlights that Applica is undoubtedly the industry leader in intelligent document comprehension.

SemEval is an ongoing series of evaluations on computational semantic systems that gathers the top scientists to compete in shared tasks every year.

ICDAR 2019 Robust Reading Challenge Task 3

 Applica’s research team recently submitted our proprietary layout-aware Language Model (LAMBERT) and data extractor for an external challenge regarding scanned and OCR receipts. Applica outperformed the majority of the teams (including some tech industry titans) to take a top spot. This confirms that our technology is incredibly well suited to extract data from receipts and invoices, as well as comprehending all types of tables, forms, and text.

More detailed information about the tasks, results, and an overview of the ICDAR challenge can be found on their website.

Trusted by companies across the world

AI-based text comprehension

Connect with an Applica expert to discuss how easy it is to automate your text-intensive business processes

Get a demo