Given the massive amounts of tabular data and information from relational databases amassed during digitalisation, it's a crucial question whether enterprises can directly construct intelligent systems from it. If possible, this approach would dramatically decrease costs, enhance data utilisation efficiency, and make data usage simpler for enterprises.
In the process of helping enterprises with customised system services, SMIRP has gained and organised extensive business knowledge from a variety of industries, including elements such as knowledge graphs, Q&As, dialogues etc.. This accumulation of knowledge offers us a competitive edge in investigating the creation of personalised knowledge graphs and dialogue-related intelligent systems for enterprises.
The potential solutions include elements like intelligent dialogue and intelligent BI analysis. SMIRP has made initial forays into this area, utilising technologies such as Text2SQL and Large-scale Table Pre-training Models. Below are several simple examples:
Natural Language to SQL
Results in Table and Line Chart
Natural Language to SQL 2
Results in Table and Bar Chart
Follow-up Question 2
Through these explorations, we anticipate making further progress in assisting our clients to construct their own intelligent systems.