Hitachi

Hitachi Develops Basic Artificial Intelligence Technologies to Strengthen Management Decision-Making at Global Enterprises

As a first step toward the multilingual use of AI systems to provide opinions, Hitachi has developed a basic artificial intelligence (AI) technology that analyzes vast volumes of Japanese text data on various issues of interest and provides opinions (both pro and con) in Japanese, along with reasons and grounds for these opinions. Applying deep learning*1 eliminates the need for dedicated programs to identify text in each language and moves closer to AI capable of supporting and strengthening management decision-making for global enterprises.

AI capable of generating logical dialogue in Japanese

With the goal of establishing AI capable of supporting enterprise decision-making, Hitachi has developed basic AI technologies that analyze vast volumes of English text data and present opinions in English.*2 The original technology required programming based on language-specific grammar rules—for example, English grammar for the English system—to identify and extract such reasons and grounds. Applying this approach to another language required developing still another dedicated program, a roadblock to expanded use of the system. Applying deep learning, the newly developed basic AI technology makes it possible to identify reasons and grounds with deep connections to the topic at issue without dependence on grammar rule descriptions. This enables the development of general-purpose systems capable of handling text data across a wide range of languages, including those in Japanese, and capable of meeting the needs of Japanese firms.

Advancing AI technologies to support enterprise decision-making

AI systems capable of logical dialogue undergirded by this basic technology automatically learn from the rules and patterns in the sentences representing reasons and grounds extracted from thousands of articles and become capable of identifying sentences that represent reasons and grounds in new articles. These systems also incorporate an attention mechanism to support deep learning to estimate which words and phrases are worthy of attention in texts like news articles and research reports. This improves performance in identifying points that require attention, including words and phrases related to specific topics and values. Through continuing research and development on this fundamental technology, Hitachi intends to advance the state of AI capable of supporting decision-making at global firms.

*1
Deep Learning: A neural network machine learning model based on the mechanism of nerve cells. The structure of a neural network is comprised of three layers: an input layer, an intermediate layer, and an output layer. In Deep Learning, the intermediate layer is increased to enable the expression of even more complex models than previously possible, achieving higher recognition rates in the field of voice and image recognition.
*2
Hitachi News Release on July 22, 2015: “Hitachi Developed Basic Artificial Intelligence Technology that Enables Logical Dialogue

Published: June 2, 2016