Re: similarity influenced ranking

Thank you Krzysztof and Riccardo. I will look into those resources more
carefully, taking time today and tomorrow. But from glancing over them, I
get the feeling that they provide ways of measuring similarity between
entities/concepts. Let me give a simple example for my problem that I am
thinking now.

Lets assume there are thee entities named, A, B, and C. They initially have
a ranking score for them individually and based on those scores, they are
ranked in the descending order as A, B, and C. But, A is more similar to C
than B and I want that to reflect in the final ranking. So that is where I
am thinking of spreading the similarity between entities to influence the
final ranking. At the end, I might get the ranking order as A, C, and B.
The thinking behind this is that, if someone selects A because it comes
higher in the original ranking, next, they may want to look at C, instead
of B.

One of my colleagues suggested to look at personalized page rank to spread
similarity between entities to influence the final ranking score. Any
thoughts along this line?

I greatly appreciate your thought (for already sent as well).

On Wed, Oct 19, 2016 at 3:31 PM, Krzysztof Janowicz <janowicz@ucsb.edu>
wrote:

> Yes, and there is, of course, also Claudia d'Amato's work.
>
>
> On 10/19/2016 11:59 AM, Riccardo Albertoni wrote:
>
> Hi,
>  if you are interested on  entity/instance similarity  you can take a look
> at SSONDE.
>
> The theoretical part is explained in
>
> Asymmetric and Context-Dependent Semantic Similarity among Ontology
> Instances
> http://pers.ge.imati.cnr.it/albertoni/PersonalPage/src/JodsX.pdf
>
> while the implemented prototype  is presented in
>
>  SSONDE: Semantic Similarity on LiNked Data Entities. MTSR 2012: 25-36
> http://link.springer.com/chapter/10.1007%2F978-3-642-35233-1_3
>
> Hope it helps.
> Best,
> Riccardo
>
> On 19 October 2016 at 20:07, Kalpa Gunaratna <kalpagunaratna@gmail.com>
> wrote:
>
>> Hi,
>>    Anybody knows a method that I can use to have things (i.e., entities,
>> property-value pairs) ranked based on similarity? For example, I have a
>> basic rank score and I want to boost this score based on similarity of
>> things to each other in the graph. PageRank is abit different to what I
>> have in mind and Similarity Flooding approach is about aligning things
>> between two graphs. Any thoughts along this line? Thank you in advance.
>>
>> --
>> Regards
>> Kalpa Gunaratna
>> *Home Page* <http://knoesis.wright.edu/researchers/kalpa/>* LinkedIn
>> <http://www.linkedin.com/in/kalpagunaratna>* *Blog
>> <http://kalpagunaratna.wordpress.com/>*
>> @Kno.e.sis <http://knoesis.org/> - Ohio Center of Excellence in
>> Knowledge-enabled Computing
>>
>
>
>
> --
> Krzysztof Janowicz
>
> Geography Department, University of California, Santa Barbara
> 4830 Ellison Hall, Santa Barbara, CA 93106-4060
>
> Email: jano@geog.ucsb.edu
> Webpage: http://geog.ucsb.edu/~jano/
> Semantic Web Journal: http://www.semantic-web-journal.net
>
>


-- 
Regards
Kalpa Gunaratna
*Home Page* <http://knoesis.wright.edu/researchers/kalpa/>* LinkedIn
<http://www.linkedin.com/in/kalpagunaratna>* *Blog
<http://kalpagunaratna.wordpress.com/>*
@Kno.e.sis <http://knoesis.org/> - Ohio Center of Excellence in
Knowledge-enabled Computing