Hi,
  Can anyone share the experience of how to implement Relevance Feedback in
Lucene?

Can someone suggest me some algorithms and papers which can help me in
building an effective Relevance Feedback system?

Thanks in advance.

Dexter.

Please check Grant Ingersoll's presentation at ApacheCon 2005.
He put out great demo programs for the relevance feedback using Lucene.

Thank you,

Koji

> -----Original Message-----
> From: varun sood [mailto:[EMAIL PROTECTED]]
> Sent: Wednesday, February 15, 2006 3:36 PM
> To: [EMAIL PROTECTED]
> Subject: Relevance Feedback Lucene+Algorithms
>
>
> Hi,
>   Can anyone share the experience of how to implement Relevance
> Feedback in
> Lucene?
>
> Can someone suggest me some algorithms and papers which can help me in
> building an effective Relevance Feedback system?
>
> Thanks in advance.
>
> Dexter.
>
---------------------------------------------------------------------

You might also want to look at that the LucQE project
(http://sourceforge.net/projects/lucene-qe/), which implement a couple
of automated relevance feedback methods including Rocchio's formula.

On 2/15/06, Koji Sekiguchi <[EMAIL PROTECTED]> wrote:
> Please check Grant Ingersoll's presentation at ApacheCon 2005.
> He put out great demo programs for the relevance feedback using Lucene.
>
> Thank you,
>
> Koji
>
> > -----Original Message-----
> > From: varun sood [mailto:[EMAIL PROTECTED]]
> > Sent: Wednesday, February 15, 2006 3:36 PM
> > To: [EMAIL PROTECTED]
> > Subject: Relevance Feedback Lucene+Algorithms
> >
> >
> > Hi,
> >   Can anyone share the experience of how to implement Relevance
> > Feedback in
> > Lucene?
> >
> > Can someone suggest me some algorithms and papers which can help me in
> > building an effective Relevance Feedback system?
> >
> > Thanks in advance.
> >
> > Dexter.
> >
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, e-mail: [EMAIL PROTECTED]
>
>
--
Dave Kor, Research Assistant
Center for Information Mining and Extraction
School of Computing
National University of Singapore.

---------------------------------------------------------------------

URL is http://www.cnlp.org/apachecon2005/

Koji Sekiguchi wrote:
> Please check Grant Ingersoll's presentation at ApacheCon 2005.
> He put out great demo programs for the relevance feedback using Lucene.
>
> Thank you,
>
> Koji
>
>  
>> -----Original Message-----
>> From: varun sood [mailto:[EMAIL PROTECTED]]
>> Sent: Wednesday, February 15, 2006 3:36 PM
>> To: [EMAIL PROTECTED]
>> Subject: Relevance Feedback Lucene+Algorithms
>>
>>
>> Hi,
>>   Can anyone share the experience of how to implement Relevance
>> Feedback in
>> Lucene?
>>
>> Can someone suggest me some algorithms and papers which can help me in
>> building an effective Relevance Feedback system?
>>
>> Thanks in advance.
>>
>> Dexter.
>>
>>    
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, e-mail: [EMAIL PROTECTED]
>
>
>  

--
-------------------------------------------------------------------
Grant Ingersoll
Sr. Software Engineer
Center for Natural Language Processing
Syracuse University
School of Information Studies
335 Hinds Hall
Syracuse, NY 13244

http://www.cnlp.org
Voice:  315-443-5484
Fax: 315-443-6886
---------------------------------------------------------------------

Hi
Thanks for replying.

I read your ppt. It is good. But the code or the basic relevance feedback is
not explained there. Actually I am not familiar with JSP, JUnit, Maven, etc.
I guess It will take me lot of time to actually discover how the things work
in demo program because I have to learn all those technologies first.
Is there any documentation or some brief notes on how Relevance Feedback has
been or could be done? I am looking on manual Relevance Feedback system.
Thanks,
Dexter
On 2/15/06, Grant Ingersoll <[EMAIL PROTECTED]> wrote:
>
> URL is http://www.cnlp.org/apachecon2005/
>
> Koji Sekiguchi wrote:
> > Please check Grant Ingersoll's presentation at ApacheCon 2005.
> > He put out great demo programs for the relevance feedback using Lucene.
> >
> > Thank you,
> >
> > Koji
> >
> >
> >> -----Original Message-----
> >> From: varun sood [mailto:[EMAIL PROTECTED]]
> >> Sent: Wednesday, February 15, 2006 3:36 PM
> >> To: [EMAIL PROTECTED]
> >> Subject: Relevance Feedback Lucene+Algorithms
> >>
> >>
> >> Hi,
> >>   Can anyone share the experience of how to implement Relevance
> >> Feedback in
> >> Lucene?
> >>
> >> Can someone suggest me some algorithms and papers which can help me in
> >> building an effective Relevance Feedback system?
> >>
> >> Thanks in advance.
> >>
> >> Dexter.
> >>
> >>
> >
> >
> > ---------------------------------------------------------------------
> > To unsubscribe, e-mail: [EMAIL PROTECTED]
> > For additional commands, e-mail: [EMAIL PROTECTED]
> >
> >
> >
>
> --
> -------------------------------------------------------------------
> Grant Ingersoll
> Sr. Software Engineer
> Center for Natural Language Processing
> Syracuse University
> School of Information Studies
> 335 Hinds Hall
> Syracuse, NY 13244
>
> http://www.cnlp.org
> Voice:  315-443-5484
> Fax: 315-443-6886
>
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: [EMAIL PROTECTED]
> For additional commands, e-mail: [EMAIL PROTECTED]
>
>

In the example code, take a look at the SearchServlet.java code and the
performFeedback and getTopTerms() methods, which demonstrate the use of
the term vectors.  It is fairly well commented.  You don't need maven,
JSP or JUnit for this.  On the indexing side, look at the TVHTMLDocument
for how to construct the term vectors.

As for how to do Rel. Feedback, you can search the mailing list archive,
there have been many discussions in the past that will offer insights
into RF in Lucene.  I also like the book "Modern Information Retrieval"
by Baeza-Yates, et.al as a text for the theory behind RF.  You may also
find the MoreLikeThis implementation (again, search this mailing list
and look in the Lucene contrib section) satisfies your needs.

Hope this helps,
Grant

varun sood wrote:
> Hi
> Thanks for replying.
>
> I read your ppt. It is good. But the code or the basic relevance feedback is
> not explained there. Actually I am not familiar with JSP, JUnit, Maven, etc.
> I guess It will take me lot of time to actually discover how the things work
> in demo program because I have to learn all those technologies first.
> Is there any documentation or some brief notes on how Relevance Feedback has
> been or could be done? I am looking on manual Relevance Feedback system.
>
>
> Thanks,
> Dexter
>
>
> On 2/15/06, Grant Ingersoll <[EMAIL PROTECTED]> wrote:
>  
>> URL is http://www.cnlp.org/apachecon2005/
>>
>> Koji Sekiguchi wrote:
>>    
>>> Please check Grant Ingersoll's presentation at ApacheCon 2005.
>>> He put out great demo programs for the relevance feedback using Lucene.
>>>
>>> Thank you,
>>>
>>> Koji
>>>
>>>
>>>      
>>>> -----Original Message-----
>>>> From: varun sood [mailto:[EMAIL PROTECTED]]
>>>> Sent: Wednesday, February 15, 2006 3:36 PM
>>>> To: [EMAIL PROTECTED]
>>>> Subject: Relevance Feedback Lucene+Algorithms
>>>>
>>>>
>>>> Hi,
>>>>   Can anyone share the experience of how to implement Relevance
>>>> Feedback in
>>>> Lucene?
>>>>
>>>> Can someone suggest me some algorithms and papers which can help me in
>>>> building an effective Relevance Feedback system?
>>>>
>>>> Thanks in advance.
>>>>
>>>> Dexter.
>>>>
>>>>
>>>>        
>>> ---------------------------------------------------------------------
>>> To unsubscribe, e-mail: [EMAIL PROTECTED]
>>> For additional commands, e-mail: [EMAIL PROTECTED]
>>>
>>>
>>>
>>>      
>> --
>> -------------------------------------------------------------------
>> Grant Ingersoll
>> Sr. Software Engineer
>> Center for Natural Language Processing
>> Syracuse University
>> School of Information Studies
>> 335 Hinds Hall
>> Syracuse, NY 13244
>>
>> http://www.cnlp.org
>> Voice:  315-443-5484
>> Fax: 315-443-6886
>>
>>
>> ---------------------------------------------------------------------
>> To unsubscribe, e-mail: [EMAIL PROTECTED]
>> For additional commands, e-mail: [EMAIL PROTECTED]
>>
>>
>>    
>
>  

--
-------------------------------------------------------------------
Grant Ingersoll
Sr. Software Engineer
Center for Natural Language Processing
Syracuse University
School of Information Studies
335 Hinds Hall
Syracuse, NY 13244

http://www.cnlp.org
Voice:  315-443-5484
Fax: 315-443-6886
---------------------------------------------------------------------