Foren

Terms facet for search in Liferay 6.2.X

Thomas Cucchietti, geändert vor 7 Jahren.

Terms facet for search in Liferay 6.2.X

New Member Beiträge: 4 Beitrittsdatum: 27.04.16 Neueste Beiträge
Hello everybody!

I would like to be able to get the most popular individual terms found in search results.
(An equivalent in ElasticSearch would be the "terms" aggregation)

From what I've seen, there is no facet dedicated for this in Liferay 6.2.X : is there some information about how to implement this by extending BaseFacet somewhere? Any documentation that could be helpful (besides this) ?

Thanks in advance,
Thomas
Avinash kumar, geändert vor 7 Jahren.

RE: Terms facet for search in Liferay 6.2.X

New Member Beiträge: 15 Beitrittsdatum: 15.12.15 Neueste Beiträge
Hi Ray,
I am facing the below mentioned issue related to the Search hook. Please provide some solution for that

I have uploaded the excel file containing data in the Document and Media portlet section . And after that I have deployed my search hook which is reading the data from excel and displaying upon search operation performed. The issue I am facing is what ever search result is displayed in the search result page i.e page.jsp , the same data is now displaying in every section of the control panel for eg. Web Content, wiki , sites etc... Please refer the screen shots attached for the issue faced.


The solution i am expecting is it should display searched data only in the search result page not any where else like it is displaying in web content , wiki.

Please respond .... Its urgent.
Thanks in Advance

Avinash
Thomas Cucchietti, geändert vor 7 Jahren.

RE: Terms facet for search in Liferay 6.2.X

New Member Beiträge: 4 Beitrittsdatum: 27.04.16 Neueste Beiträge
Thanks Ray for your answer.

I've ve been rethinking about that, and it could be possible to get most popular terms by faceting on a specific field where analysis has split values into individual terms.

I was faceting on assetCategoryIds field, which is indexed as a StrField in Solr, but this analysis process keeps the whole value as a single term. Maybe that if I prepare a facet targeting a field indexed with a Whitespace or Standard tokenizer, I'll be able to achieve the wanted behavior.