Benjamin, for small scale and fast prototyping, you can use Neoclipse, http://wiki.neo4j.org/content/Neoclipse_Guide which we used in this case.
For larger experiments, the JRuby bindings are quite nice, and then you can visualize then by pointing Neoclipse to the Neo4j storage files. HTH /peter neubauer Neo Technology GTalk: neubauer.peter Skype peter.neubauer Phone +46 704 106975 LinkedIn http://www.linkedin.com/in/neubauer Twitter http://twitter.com/peterneubauer http://www.neo4j.org - Relationships count. http://www.linkedprocess.org - Distributed computing on LinkedData scale On Wed, Nov 11, 2009 at 3:05 PM, Benjamin Dageroth <[email protected]> wrote: > Hi, > > thanks for the fast answers and even drawing a domain model! What did you use > for modeling? I have something similar on paper by now, perhaps I can > contribute the stuff as well and I'll try to talk my CTO into some > experimenting. > > The road to aggregation is unfortunately blocked for us, as we want our > customers to be able to drill down to really small groups and even single > visits - for this we need the raw data. Once aggregated you cannot > distinguish anymore and not answer all questions depending on the filters > that are used on the data. E.g. What are the favourite paths of Visitors from > Germany. Since this capability is one of the major sellingpoints of our > solution we'd probably try to use both databases. The relational data for all > kind of filtering and the graph database for path analysis. > > Graph Databases are really quite new yet, yes? I am trying to find some > literature on them, but Oreilly seem to have nothing yet on the topic. > > Other than often traveled paths I am by now thinking of using the database > for recommendation engines. Currently we use pretty much a brute force > approach to find out which articles have often been purchased together, but > with a graph database one should be able to implement algorithms which are > way more efficient and using information stored in the edges should also make > it possible to increase the quality of recommendations as more factors than > just the edge can be taken into account. And of course realtime > recommendations should be possible independent of nightly batch jobs. > > Thanks again for the answers, > > Benjamin > > _______________________________________ > Benjamin Dageroth, Business Development Manager > Webtrekk GmbH > Boxhagener Str. 76-78, 10245 Berlin > fon 030 - 755 415 - 360 > fax 030 - 755 415 - 100 > [email protected] > http://www.webtrekk.com > Amtsgericht Berlin, HRB 93435 B > Geschäftsführer Christian Sauer > > > _______________________________________ > > > -----Ursprüngliche Nachricht----- > Von: [email protected] [mailto:[email protected]] Im > Auftrag von Johan Svensson > Gesendet: Mittwoch, 11. November 2009 14:39 > An: Neo user discussions > Betreff: Re: [Neo] Web Analytics Use Case > > On Wed, Nov 11, 2009 at 11:02 AM, Benjamin Dageroth > <[email protected]> wrote: >> Hi, >> ... >> >> Then the best solution for this seems to me that I would run two databases: >> First I ask our current current database to give me all Visitors, to whom >> this criteria applies, and then ask neo4j to look only at the edges of these >> visitors. Or would neo4J be powerful enough to deliever a similar >> performance as traditional RDBMS Systems when confronted with data that is >> not really resembling a graph? Is it usually easy to transform a traditional >> schema into a graphDB Schema that performs just as well? >> > > Hi, > > Running two databases (graph db + RDMS) may be good idea in this case. > RDMS are really good at working with set operations on the full > dataset. On the other hand it is often possible to calculate/aggregate > such information on the fly when data is updated in a graph db. > > Regarding transforming a traditional schema into a graph I would say > it is easy. I usually recommend not to look at the schema, instead > look at the domain and its data and transform that into a graph. > People are playing with automatic relational schema -> graph db > transformation tools. Have a look at Peters SQL importer here: > http://wiki.neo4j.org/content/SQL_Importer > > Regards, > -Johan > _______________________________________________ > Neo mailing list > [email protected] > https://lists.neo4j.org/mailman/listinfo/user > _______________________________________________ > Neo mailing list > [email protected] > https://lists.neo4j.org/mailman/listinfo/user > _______________________________________________ Neo mailing list [email protected] https://lists.neo4j.org/mailman/listinfo/user
