Machine reading comprehension engine MRC
The machine reading comprehension engine goes through related documents to find an answer to the question and finds the answer. Especially, AI Suite’s machine reading comprehension engine compensates for the shortcomings of the existing technology, in which documents were fed to find answers to questions, by using various knowledge resources owned by Saltlux. Thus, it presents an open QA type service in which answers are found without document input. Or, it can be used to extract desired information from a document.
- Providing optimal machine reading comprehension results by situation
Multi-learning models are references to AI Suite’s machine reading comprehension engine. You can choose the best learning model that can vary from question to question, which is much better than other machine-reading comprehension engines that rely on a single learning model.
- Can be used as Open QA
Saltlux uses a variety of knowledge resources collected over the past 20 years in its machine reading comprehension engine for QA. By automatically extracting a document or paragraph from a knowledge resource that it already has, and entering it into a machine reading comprehension engine, a user can enter a question without entering a document and obtain an answer, thus improving the Q&A service.
< Machine reading comprehension through automatic document search/input >
- Service expansion in connection with AI Suite engines
The machine reading comprehension engine can enhance its usefulness by linking with other engines in AI Suite. Simple Q&A services are possible during dialogue in connection with the dialogue processing engine, and in conjunction with the Q&A technology, it provides an ensemble of deep Q/A, thus improving the quality of answers. In addition, the machine reading comprehension engine can be used when the knowledge automatic extraction engine extracts knowledge from structured/unstructured documents or table data.
Main features and specifications
- Machine reading comprehension service
The machine reading comprehension engine does not currently provide a separate management tool and is provided as a system installation type. The API that can access the machine reading comprehension service operates when the system is installed and run, and users can use the API to implement various services with the engine.
< Machine reading comprehension test and visualization screen >
- Learning data management
It provides a feature in which learning/evaluation data can be created and managed for machine learning. Machine reading comprehension learning data should be constructed as a pair of fingerprints, questions, and answers. Especially, it is not easy to construct learning data because data should be constructed that includes the index information located in the corresponding fingerprint. The learning data management tool provided by the machine reading comprehension engine can easily construct data because it generates questions about documents required for learning and evaluation, and selects appropriate answers by a drag-and-drop method.
- Learning Management
It provides a feature in which the process of learning the model for machine reading comprehension can be performed in the web management tool. UI is provided to help intuitively understand and manage the workflow such as selecting a data set for learning, setting learning options, and evaluating learning results. You can create a learning model by selecting your own learning algorithm and setting the parameter values in it. Learning progress can be visualized and checked, and learning history is also managed.
- Model validation and distribution
The trained model should be reflected in the machine-reading comprehension engine in service. It provides a feature in which target service applications to deploy a model are registered and managed, and models can be distributed, backed up, or cancelled with a test or management server.