Hi Tibor,

On Wed, Oct 6, 2010 at 10:39 PM, Tibor Simko <[email protected]> wrote:
> On Wed, 06 Oct 2010, Roman Chyla wrote:
>> doc: 10
>>   cited: 3,6,80,90,89...
>>   citing_author: witten, frank, lagra, ngeyen, chu, thuey...
>
> What citing_author holds?  An equivalent to data points for
> citedby:author:lagra?  Not practical to store stuff like that next to

It was trading space for speed, why that is not practical? Especially
if the values are data points

> every record.  Note that citedby/refersto operators can operate on any
> query, not only on authors.  See my `refersto:keyword:muon' example.
>
>> the lucene query with the same effect then is:
>>
>> ((author:ellis +citedby:witten -author:witten) +keyword:muon) -->
>> cluster_by(len(cited))
>
> Not quite.  `refersto:keyword:muon' gives you a set of papers that cite
> some paper from the set of papers that are tagged with keyword muon.
> Gives 93k hits on INSPIRE.  To be compared to the set of papers tagged
> with keyword muon, 22k hits on INSPIRE.

OK, I see - that is a very nice case and I understand little bit more
- could you point me to some info about how this 2nd order lookup is
done, or where int the code, or discuss it here? I guess it will apply
also for the chain of 2nd order lookups

>
>> notes:
>>  - citedby:author:witten -- it doesnt make sense to me that it could
>> be sb else than other author
>
> Note sure what you mean.  Gives collection of all papers that are cited
> by any of the paper written by Witten.  It could be any author,
> including Witten itself (think self-cites).  Try it on INSPIRE.

I meant this case, the toy-index would allow that query...

>
>>  - 2nd order links must be carefully prepared (but honestly, how many
>> of those 2nd order relations are really needed, and really used? this
>> number is probably low...)
>
> Cite summary, co-cited with, etc are all second-order operations.

I obviously don't know most of the 2nd order operations, but musing
about how they are implemented - if somebody could point out reasons
why they are not possible or very difficult to implement with the
search engine index, that would help me to understand much more.

Cheers,

  roman



>
> Best regards
> --
> Tibor Simko
>

Reply via email to