titan - Gremlin - Giraph - GraphX ? On TitanDb -

I need some help to confirm my choice ... to know if you have any information I can give my storage database with a tinted cassandra. I have a very big graph. My goal is to use Amilib later on the graph.

My first thought: use Titan with graphics, but I did not find anything or development did not progress ... Tinkerpop is not ready yet. So I have a look for the Giraffe. TinkerPop can talk to Tinker Pop from Titan Rector.

My question is: What is the use of the Giraffe? Gremlin looks and is distributed like this.

Thank you very much for explaining me I think that I do not really understand the difference between Gremlin and Giraf (or GraphX).

There is a good day.

Interesting questions I'm on the same track.

Your question about MLBB first, I think that means you mean, the implementation of machine learning (ML) at the top of Apache Spark is my conclusion: using data in your own / based graph database For purposes such as clustering and classification, you want to run ML algorithms. Please note that you can use graph processing algorithms such as page rank described by Spidi, which do things like clustering on your Titan / Cassandra graph database. In other words: Clustering is not required when your starting point graph is database.

Apache Spark MLBB is proof of future and it is widely supported, their most recent announcements were about the new ML algorithm, however, another Apache ML project, about the amount of supported ML algorithms Is more mature in Apache Mahto has adopted Apache Spark as its data storage layer, so I mentioned this in this post. In addition to Apache Spark Off-Memory Computing, the machine offers MLLIB mentioned for learning, which is like Spark, which is a graph processing system, as explained by Spidi and is for streaming data processing.

I Consider Apache as a logical data layer, presented on top of storage layers such as cassandra, headop / hcb and hbb as RDD (flexible distributed datasets). Apache Spark provides a connector to the cassandra. Note that RDDs are immutable, you can not change the data using Spark, you can only process and analyze the data in Spark. Regarding Apache Spark Logical Storage Layer RDD: You can compare an RDD in the good old SQL Bar, for example, the RDD offers you a view, for example Cessandra HBbase has a table, keep in mind that Apache Spark provides APIs for 3 development environments: Scala, Java and Python.

Apache is a functional equivalent for Spark Graphics, a graph processing toolset too. Apache Girpa uses Hadop as a storage layer. You are using Titan / Cassandra, when you select Apache kernel as your solution, you will probably enter data migration tasks. Secondly, you have started your post with a question about ML using MLLB and Apache Girpa. MAL is not a solution.

Your findings about Grapher and Gramme are not correct: although they are not using both a graph database. Graff is a solution for graph processing as Spidi explained. Using graph, you can perform graph analysis algorithms such as page rank, e.g. Curriculum to the graph database using complex relationships (edges) between entities (parentheses) that get results of

result result for most of whom are most followers, while for traversing.