Titan graph database issue with bulk loading for any datasource

Question

I am very new to Titan Graph database. Please let me know can we perform bulk loading in titan db using Batch Graph for any format of files(example: pdf, doc, xml, xls, csv and etc). Kindly, please reply if anyone knows about it. Thanks in advance. Regards, Kamal Verma

Answer 1

If you are new to Titan, you should not be looking at BatchGraph for data loading because that is a Titan 0.5.x (and TinkerPop 2.x) concept. If you are "new" you should be looking at Titan 1.0 (and TinkerPop 3.x):

http://s3.thinkaurelius.com/docs/titan/1.0.0/

http://tinkerpop.apache.org/docs/3.1.0-incubating/

Titan takes its bulk loading cues from TinkerPop which utilizes the BulkLoaderVertexProgram also known as BLVP

BLVP and ScriptInputFormat provide you all the flexibility you need to load line delimited data (i.e. for csv, json, xml, etc each line is a vertex). That obviously won't help you with data held in pdfs and word documents. You would have to pre-process those into line-delimited format to be able to load them in this fashion.

I highly recommend that you focus on learning TinkerPop before getting too deep into Titan. TinkerPop is a large ecosystem in and of itself and having a solid basis in that will make it possible to understand the nuances of Titan. I would suggest you start by reading the TinkerPop Getting Started Tutorial before progressing further.

For what it is worth, if you are stuck on Titan 0.5.x for some reason, you can use BatchGraph for loading all the data formats you mentioned, but it does not offer a method to do so natively. You would have to write a script to parse each of those formats and convert it to vertices/edges that would be loaded through BatchGraph.