The Electronics and Telecommunications Research Institute (ETRI) is expected to be able to effectively use AI contact centers, educational institutions, etc. to easily explain specialized knowledge through the development of artificial intelligence (AI) that can assist with decision-making by experts in law, consulting, etc.

ETRI announced on the 19th that it has developed an artificial intelligence technology that supports explainable expert decision-making.

The technology developed by ETRI is an explainable self-verification-based augmented search generation (RAG) technology that presents the appropriateness and usability of search evidence and the appropriateness of the correct answer together with the correct answer through self-verification.

It can also ensure the reliability of the generated answers and provide explainability.

This technology is expected to be used in legal advice assistance systems, various consulting systems, intelligent customer support services, and corporate knowledge management systems.

The research team explained that this technology, developed based on their own Korean-optimized tokenized language understanding model (MoBERT) and neural paragraph retrieval technology (MV-ColBERT), provides much more accurate and faster answers in specific fields than existing AI-based conversational chatbots.

AI technology that supports legal advice searches and analyzes related case law and laws when a lawyer, legal practitioner, or client inputs legal-related content. Infer legal acts.

It allows you to easily interpret and verify complex case law in everyday terms, thereby providing reliable advice in the legal field.

Counseling support AI technology has the ability to classify and summarize counseling conversations with customers and explain the reasons. It analyzes and summarizes counseling conversations from various perspectives and provides appropriate answers.

This technology was developed in collaboration with F&U Credit Information Co., Ltd., and is currently being commercialized through technology transfer.

ETRI Language Intelligence Lab Director Kwon Oh-wook said, “In a reality where AI-generated results are difficult to understand and difficult to determine whether they are true or not, explainable decision support technology will be of great help in professional consulting fields such as law and finance.”

Dr. Baek-Man Bae, the project manager, said, “We plan to successfully complete the PoC and create a good application case of expert-supporting AI technology to expand it to various specialized fields.”

This technology development is being carried out as part of the 'Human-Centered Artificial Intelligence Core Source Technology Development Project' supported by the Ministry of Science and ICT and the National IT Industry Promotion Agency, and the research team plans to expand the application of the technology to various public service fields in the future.