Roadmap for Lucene expansion, maybe?

[scott molinari]

I just took a look at a this presentation 

http://www.slideshare.net/wolf4ood/orientdb-lucene

and one of the requirements on our system will be a search engine as a service. If we can get that integrated in the same database, that lowers complication of our system a lot.

Currently, the Lucene index is ok, but not close to or on par with the likes of Elasticsearch and that is simply because the Orientdb API to use Lucene indexes is still fairly rudimentary. 

Is there a clear -> This will happen in these versions with the Lucene index API roadmap? 

Or can Luca or Enrico possibly share the views of what is planned, but not clearly communicated yet? 

I was looking at Titan and their ES integration looks very interesting. I am sure ODB could get closer to this kind of search engine type of capability, by exposing more of the Lucene's capabilities. The slide presentation also hints at some added, but fairly basic search engine needs and just even knowing this is going to happen for sure would help us out.

Thanks! 

Scott  

[scott molinari]

A friendly bump. 

Scott

[Riccardo Tasso]

Hi, I am also interested since my application uses both orientdb and lucene but can't use orientdb-lucene because it offers poor functionalities.

Cheers,

   Riccardo

2015-09-01 8:51 GMT+02:00 scott molinari <scottam...@googlemail.com>:

A friendly bump. 

Scott

--

---
You received this message because you are subscribed to the Google Groups "OrientDB" group.
To unsubscribe from this group and stop receiving emails from it, send an email to orient-databa...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[scott molinari]

To put some perspective on this. Any decent website wanting to publish content or any application with any good bit of text within their application, will absolutely need a very good full text search. It is why Elastic Search has so much success. I've worked with a ton of websites and as soon as they get to any bit of size and traction, the first thing they look for is a faster, better and more comprehensive search system. Our first contemplation was to build Elastic Search next to OrientDB, but that means more effort, more money (which the customer must pay for too), two different systems and tons more complication. 

So, putting 1 + 1 together, there is a general and very real need for a very feature rich full text index and query system in any data storage. If a single database can do it, then that makes that database very, very attractive.

In other words, if a database has....

Powerful Graph Queries,

Powerful Document Queries, 

and Powerful Full Text and Spacial Queries

Then that would be just.....Wow!

A side question, but could it be possible to index with Lucene through Orient, but query against the Lucene index through Solr?

Scott  

[Enrico Risa]

Hi Scott

The roadmap of OrientDB 2.2 is written here

http://orientdb.com/docs/2.1/Roadmap.html#release-22.

About Lucene I can tell you that we are rewriting the spatial module, that will be more advanced:

Some of the features are

-Indexing not only points (Polygon,Line..etc).

-New Spatial Operators

-New Spatial Functions.

Then we have schedule for OrientDB lucene the support for Facet Searching, but i think

this will land in 3.0

For supporting Lucene through Solr i don't know if it is possible.

Cheers 

Enrico

--

[scott molinari]

That sounds great. How far off is 3.0 approximately? 

I've copied a list of main features in Lucene and I've ordered them by priority, as I'd like to see them happen in ODB. It is just a suggestion. Please let me know, if OrientDB offers any of these features already. I added a couple of comments too (->). If I need to start that as an issue, please let me know. 

• many powerful query types: phrase queries, wildcard queries, proximity queries, range queries and more -> in other words, take full advantage of the query engine
• ranked searching -- best results returned first
• multiple-index searching with merged results -> the merged results aren't terribly important, but a nice to have
• fielded searching (e.g. title, author, contents)
• flexible faceting, highlighting, joins and result grouping -> mainly important for ecommerce. 
• sorting by any field
• allow simultaneous update and searching
• fast, memory-efficient and typo-tolerant suggesters
• pluggable ranking models, including the Vector Space Model and Okapi BM25

Scott

[scott molinari]

Ok. Got my answer about when 3.0 is possibly coming out. End of the year or beginning of next. That is cool and I hope it pans out. 

Scott

[Roberto Franchini]

On Tue, Sep 1, 2015 at 1:57 PM, scott molinari
<scottam...@googlemail.com> wrote:
[cut]

> many powerful query types: phrase queries, wildcard queries, proximity
> queries, range queries and more -> in other words, take full advantage of
> the query engine
> ranked searching -- best results returned first
> multiple-index searching with merged results -> the merged results aren't
> terribly important, but a nice to have
> fielded searching (e.g. title, author, contents)
> flexible faceting, highlighting, joins and result grouping -> mainly
> important for ecommerce.
> sorting by any field
> allow simultaneous update and searching
> fast, memory-efficient and typo-tolerant suggesters
> pluggable ranking models, including the Vector Space Model and Okapi BM25

Hi to all.
I invite you to open issues for each feature request on
https://github.com/orientechnologies/orientdb-lucene

Thanks,
FRANK

[scott molinari]

Thanks Frank. I added a few.

Scott