Knowledge graph data analysis for organizational intelligence to support continuous business improvement
A Talk by Yuta Yajima , Shigenori Matumoto and Kunihiko Harada
About this Talk
The collaboration between management and frontline staff has long been a critical issue in corporate management. However, it is challenging to clarify or model the relationship between the issues faced by management and those faced by frontline staff. Additionally, knowledge within the organization tends to be confined to specific departments or individuals, making it difficult to share, leading to information shortages and redundancies, which hinder efficient decision-making and innovation.
In response to these problems, we have a track record of business achievements across various industry fields and possess cross-sectional problem-solving know-how. Therefore, we believe that leveraging digital knowledge utilization presents a business opportunity.
Currently, generative AI is increasingly being used to enhance and automate existing operations. In the near future, it is expected to be utilized for advanced decision-making in management and the transformation of business processes. As one of the few companies that can combine AI technology with expertise in corporate management, operations, and products, we aim to be at the forefront of providing these technologies and solutions to the world.
The target audience includes data scientists, business analysts, and sales professionals who face challenges with the speed of solving customer issues. Additionally, it also targets engineers interested in applications of knowledge graphs and GNNs. Participants will learn methods for data-driven decision-making, how to apply past success stories to new business challenges, and the technical implementation of these methods.
We extract and organize customer information into four graph data structures, and utilize it effectively. The four graph data structures are as follows:
- Value Chain: Share policies based on the flow of value among stakeholders, which is crucial for expanding sales and horizontal deployment.
- Value Graph: Set goals based on the causality of business issues and operational measures, which is important for verifying the effectiveness of analysis.
- Profiling: Identify operational constraints and short-term effective measures, which are important for delivery realization.
- Solution: Select appropriate analytical methods based on the types of analysis and data analysis constraints, which are important for systematization and implementation.
The aim is to leverage the information accumulated through projects with customers across various industries to generate new ideas in new customer projects by utilizing cross-industry knowledge. To achieve this, we are developing technology that uses GNN (Graph Neural Networks) to extract and reconstruct graph structures from previously created graph information, leading to the creation of new challenges.
Furthermore, to create these graphs for new customers, we are also developing technology that automatically generates initial graphs by utilizing graph construction know-how and generative AI.
We help organize and visualize customer information in a graph structure to support business reform in companies by facilitating consensus building and decision-making.
We’re a cloud tech company that provides organisations around the world with computing infrastructure and software to help them innovate, unlock efficiencies and become more effective. We also created the world’s first – and only – autonomous database to help organise and secure our customers’ data.
