Building a Real-time Recommendation Engine With Neo4j - Part 1/4 - William Lyon - OSCON 2017

Поделиться
HTML-код
  • Опубликовано: 9 янв 2025

Комментарии • 9

  • @VinayNarayanaK
    @VinayNarayanaK 6 лет назад

    Hey William,
    The cypher Query at 27:40 Considering the direction of relationship
    Is it *Match (t)

  • @sercanyildirimtugcann
    @sercanyildirimtugcann 2 года назад

    Hi William, I am too confused about Neo4j and wanted to ask you directly. Any help would be appreciated. What I want to create is search engine for documents ok. I am also interested in grap machine learning and recommendation engines but also want my platform seem a bit interactive and stylish so I will use react js.
    The question is that which way is much more compatible with Neo4j? Node. js+react or Django+react? What do you suggest me as beginner on graph databases? I want to fully dive into this topics.
    Thanks in advance!!!

  • @anicethounkpe582
    @anicethounkpe582 6 лет назад +2

    Hey William Lyon! Thanks for your presentation.
    The link for your slides is not available anymore.

    • @lyonwj
      @lyonwj  5 лет назад +2

      Sorry about that - looks like the dropbox link died :-( I've added the slides to the Github repo here: github.com/johnymontana/oscon-neo4j-recommendations

  • @PriyaAggarwal24
    @PriyaAggarwal24 5 лет назад

    Hello Will, Great session!! :) I just have one small question if you help answer that please. You mentioned that there could be multiple tags associated with a node. Could you please share an example where having multiple labels for a node would be useful?

    • @dhananjayghanwat
      @dhananjayghanwat 4 года назад

      Let's say you have a Service Node which represents every service (train) in a plan. Now in a typical plan for 6 months, you might have hundreds of thousands of services which run on different days. Now instead of having just one label "Service", you can also label them as SundayService, MondayService and so on to represent Services which run on a given weekday. By doing that you will have the total number of SundayService nodes far less than Sevices nodes. So now you can use label SundayService in your cypher queries to work on smaller data set rather than huge dataset of Service nodes, which will make you query much faster.
      Hope it helps

  • @krishnachaitu17
    @krishnachaitu17 3 года назад

    will ,great session thank you for it

  • @diarmuidbrady8987
    @diarmuidbrady8987 3 года назад

    Where can I get the data for this if it is not on the sandbox

    • @rickyS-D76
      @rickyS-D76 2 года назад

      Here in his github link: github.com/johnymontana/oscon-neo4j-recommendations/blob/master/graph-based-rec-sys.ipynb