Models overview
WorkFusion offers models for but not limited to the following use cases:
Classification
Classification stands for arrangement in groups or categories according to established criteria. Both binary and multi-class are supported.
Examples include email classification, document identification, classification of pages within documents, approval process.
Information extraction
Information extraction involves extracting structured information, or key facts, from unstructured or semi-structured documents.
Examples are extraction of details from invoices, claims, and financial documents.
Hyper model
There are two types of hyper models:
A fixed set model containing an optimal model configuration created by a Data Scientist for a specific use case.
A Search Engine model that finds the optimal configuration by running a Search Engine. Typically requires more time and resources compared to a fixed set model.
Each hyper model has a number of parameters defining prerequisites. For example, minDocs stands for the minimum amount of documents required for training,deltaDocs is the number of documents that can trigger automatic re-training.
Single model vs. multi-model
Under the hood, WorkFusion allows two types of models for handling complex documents:
- single model
- multi-model
Single model approach
This approach implies training a single binary Model for all fields. It works when you need to extract fields with the same semantic meaning. All fields are equally important and used in a use case. Fields are subject to post-processing.
This approach is optimized for training and execution.
Multi-model approach
This approach implies training a set of individual binary models per field. It works when the semantic meaning of fields can vary, which means different subsets of fields are used in a use case.
This approach is optimized for a higher automation rate and accuracy.