Why graphs are important for social PLM strategy?

Why graphs are important for social PLM strategy?


I want to continue my series of thoughts about PLM and social technologies. If you missed the beginning, you can catch up here – It is not easy to add social to PLM. The topic I want to bring today is graphs. The topic of graphs became popular for the last few years as technology moved into social networking and social data analytics. We can see a growing amount of examples while graph technologies are successfully used to boost social strategies.

I’ve been touching graph earlier on my blog – Social PLM, graphs and organization overlap. Large companies are looking into more use cases. One of them is to leverage graph for collaboration. Here is an example from Microsoft – Oslo and Office Graph: new technology in data discovery and search. Another example is Facebook graph search. Last, but not least – Google Knowledge graph.

However, graphs are not limited to Google, Facebook, Microsoft and few other large companies. Graph technologies are getting more popular and we can see companies applying them to variety of scenarios related to analysis of connected information.

Medium article – How Medium Goes Social speaks about how Medium publishing website answering on typical questions like – Have any of my friends liked this post? Can I tell all my friends once I’ve published something new? How can I find new people to follow? The picture below shows example of graph information. Medium is using Neo4j to capture and query graph data.


The following video shows how to develop social recommendation algorithm using Neo4j and cypher language.

All examples together made me think about usage of graph technologies to boost a specific “social” strategy for PLM. In my presentation last year – PLM and Data Management in 21st century, I’ve been demonstrating variety of modern data management technologies and how they can be used in different PLM applications. It looks like graphs can work well for social and provide real benefits. Graph databases are getting popular. Wikipedia article about Graph databases contains 20+ different names. It is maybe a time for PLM architects to have a look.

What is my conclusion? Graphs and social are coming together.  In my view, low value proposition was the main reason behind  failures of social technology application in PLM domain in the past. Rethinking value proposition is important. To combine social value proposition with efficient technology can be a way to bring some interesting new social features into PLM space. Just my thoughts…

Best, Oleg

photo credit: Matthew Burpee via photopin cc


Share This Post

  • Would you limit the usage of graphdb for social? or would it be also a good tool to support the entire PLM?

  • beyondplm

    Yoann, thanks for that clarification. The article was about ‘social’ topic. Of course, I’m not limiting usage of graphdb to social tech only. The link to my presentation about Database tech and PLM speaks about it in a more generic way. GraphDB tech has some advantages and disadvantages and in my view, part of database landscape that can be used.

  • Pingback: From the Community: December - Neo4j Graph Database()

  • Pingback: From the Community: December 2014 - Neo4j Graph Database()

  • Pingback: From the Neo4j Community: Best of December 2014()

  • Great article. Linkurious, the young graph visualization startup I work with, is seeing a lot of traction for PLM. I hope you’ll write more articles on the subject !


  • beyondplm

    Jean, thanks for your comment and link!

  • Sorry Yoann/(andOleg) I only just saw your post – but I think we are thinking along the same lines. Obviously initially graph DBs would perform poorly or normal sized data sets but for the right industry other benefits may out weigh raw power. I guess I’m thinking Christenson’s Innovator’s Dilemma examples esp on Disk Drives.

  • beyondplm

    Paul, thanks for your comment! I agree, Graph databases today cannot provide a fully functioning backend for PLM system as RDBMS because of the maturity level. However, the dynamics in databases development are very high these days. I put couple of slides last year speaking about Data Management in PLM for 21st century. You might take a look here — http://beyondplm.com/2013/10/25/tech-soft-3d-techtalk-plm-and-data-management-in-21st-century/

  • Thanks Oleg – it seems like my original post didn’t make it to the thread. (generally I was referring to Neo4J) Your slides are very useful and I concur on graph databases & the potential use for product data. I believe that databases such as Neo4J are also useful for evolving product data at the Fuzzy front end of product development by being able to go from loosely linked ‘Ideas’ and concepts and then being refined to more concrete product structures by refining the ‘relationships’ as well as the Component ‘nodes’.

  • Hi Paul, I’m actually working on some solution using graph (can’t tell more for now) and the small frustration I have so far is that the advantages of using such database is not seen by the users except that I can adapt to customer’s need much faster as I don’t have a schema (allowing a more lean product development process). But overall, it’s fantastic to work on BOMs using graphs, version items that are handled by an ECO, cancel an ECO,… I can see the biggest value for developers and administrators.

  • beyondplm

    Hi Paul,

    Re your comment. I was trying to see maybe on of your comment didn’t pass spam filter. But, I cannot see anything there. So, just send me your comment on oleg at beyondplm dot com- I will post it.

    Yes, there is a potential to use Graph databases to work on product data. To use graphs in social domain is more straightforward and what is important can be complementary, which won’t raise an issue of comparison between graph databases and other data management technologies.

    Thanks for your comments!
    Best, Oleg

  • No worries Oleg – I or my fingers probably messed up – thanks for looking tho’