PLM is all about bringing right information to the right people at the right time. Nice marketing you can say. And it sound simple… Huh? Well, it is sounds very simple, but implementation of this simple formula is not very simple. It takes time and effort to bring right data from PLM databases. Sometimes, the data is coded in a very complex set of tables and relationships.
How to get data out to users when you need it the most. For the last few years, the idea of applying search technology was promising. PLM products have adopted variety of search implementation – full text, graph, 3D, semantic etc. Even search is one of the most promising strategy to democratize the data extraction, it is not an easy way to do so. Queries can bring too many results and to apply filters can be sub-optimal.
What is the alternative to search? Structural queries are much easier to implement and they can return more precise results. However, to implement such queries wasn’t a simple thing. But, there are some good news coming on the horizon. My attention was caught by Tech Crunch article – Salesforce is using AI to democratize SQL so anyone can query databases in natural language (http://techcrunch.com/2017/08/29/salesforce-is-using-ai-to-democratize-sql-so-anyone-can-query-databases-in-natural-language/). Read the article and draw your opinion.
The article speaks about the development of Seq2SQL – a technique Salesforce.com is using to build more faster and comprehensive queries based on natural language analysis. The following passage can give you an idea of what is that and how it works.
Their recent paper, Seq2SQL: Generating Structured Queries from Natural Language using Reinforcement Learning, builds on sequence to sequence models typically employed in machine translation. A reinforcement learning twist allowed the team to obtain promising results translating natural language database queries into SQL.
In practice this means that you could simply ask who the winningest team in college football is and an appropriate database could be automatically queried to tell you that it is in fact the University of Michigan.
“We don’t actually have just one way of writing a query the correct way,” Victor Zhong, one of the Salesforce researchers who worked on the project, explained to me in an interview. “If I give a natural language question, there might be two or three ways to write the query. We use reinforcement learning to encourage use of queries that obtain same result.”
Here is an interesting video shows you the idea how it might work.
What is my conclusion? The idea of generating SQL can be potentially applied to complex PLM databases struggling how to extract data in a meaningful way. Query generated using the algorithm described in the article and it is promising. So, the time when you will be query your PLM database using NL tools is not so far. Just my thoughts…
Want to learn more about PLM? Check out my new PLM Book website.
Disclaimer: I’m co-founder and CEO of openBoM developing cloud based bill of materials and inventory management tool for manufacturing companies, hardware startups and supply chain. My opinion can be unintentionally biased.