[Freebase-discuss] How to find an apple in Suggest/autocomplete ?

[Jeff Prucher]

It turns out to be more complicated than that, and it's not really
possible to exclude types from the blacklist as we had thought. So
we'll need to remove the SKOS Concept type from the affected
instances, which means we'll first need some schema to hold the
broader/narrower ingredient stuff that Mike added. Which I'll put in a
new thread.

Jeff

On Tue, Mar 15, 2011 at 2:00 PM, Gordon Mackenzie <gmackenz at google.com> wrote:
> Well, Thad's getting closer, but to save all our heads from exploding...SKOS.
>
> We specifically blacklist topics typed as SKOS concepts in the
> property suggest (excluding from the blacklist filter topics co-typed
> as book subject). As a work-around until we find a better solution for
> using freebase topics as vocabulary, we will add a few more types to
> be in the exclusion list for this blacklist filter which should help
> with revealing in property suggest a large portion of the 500+ topics
> (with multiple co-types) currently typed as SKOS concepts. Malus
> domestica will live again.
>
> In the short term, do as Tom discovered: use the /m ID of the topic.
>
> On Tue, Mar 15, 2011 at 1:29 PM, Thad Guidry <thadguidry at gmail.com> wrote:
>> Hmm...
>>
>> {"field":"type_strict","value":"a"},"message":"Value must be one of:
>> any; should; all}
>>
>> To me it appears that with "should" it is not being provided a strong
>> enough signal that Apple is a "organism classification".
>>
>> versus with type_strict=all that Apple appears.  topicedit.js seems to
>> contain the necessary code to change.
>>
>> Or perhaps it's just the Suggest hinting and scoring that needs a tweak here ?
>>
>> On Tue, Mar 15, 2011 at 3:08 PM, Thad Guidry <thadguidry at gmail.com> wrote:
>>> Yeah, it's something regarding type_strict being added, and without
>>> that seems to give better results:
>>>
>>> http://www.freebase.com/private/suggest?prefix=apple&type=%2Fbiology%2Forganism_classification&category=instance&all_types=false
>>>
>>> Hope this helps someone fix it up.
>>>
>>> On Tue, Mar 15, 2011 at 3:04 PM, Thad Guidry <thadguidry at gmail.com> wrote:
>>>> Here's the reason, it's using this XHR:
>>>>
>>>> http://www.freebase.com/private/suggest?prefix=apple&type=%2Fbiology%2Forganism_classification&type_strict=should&category=instance&all_types=false
>>>>
>>>> Type_strict=should perhaps is getting in the way ??
>>>>
>>>> On Tue, Mar 15, 2011 at 2:37 PM, Jeff Prucher <jprucher at google.com> wrote:
>>>>> On Tue, Mar 15, 2011 at 8:01 AM, Tom Morris <tfmorris at gmail.com> wrote:
>>>>>> On Mon, Mar 14, 2011 at 5:51 PM, Gordon Mackenzie <gmackenz at google.com> wrote:
>>>>>>> On Mon, Mar 14, 2011 at 2:39 PM, Thad Guidry <thadguidry at gmail.com> wrote:
>>>>>>>> Tried just now with apple and was the 2nd one on the Suggest drop down.
>>>>>>> What Tom is mentioning is that in the 'Higher classification' property
>>>>>>> field of 'Organism Classification' type the suggest is not returning
>>>>>>> "Apple" and I wonder if it's something unique about that particular
>>>>>>> topic, as other topics that are classified as 'Organism
>>>>>>> Classification' are appearing in the suggest field as expected.
>>>>>>
>>>>>> What Gordon said.
>>>>>>
>>>>>> It's almost as if it's being explicitly blacklisted somehow in that
>>>>>> specific usage.
>>>>>
>>>>> It's true for oranges, too
>>>>> (http://www.freebase.com/view/biology/itis/28889). I thought maybe it
>>>>> had something to do with being an ingredient, but basil seems to work
>>>>> fine.
>>>>>
>>>>> Jeff
>>>>> _______________________________________________
>>>>> You are receiving this message because you are subscribed to the Freebase-discuss mailing list.
>>>>> To post a message to the list: Freebase-discuss at freebase.com
>>>>> To unsubscribe, view archives, etc: http://lists.freebase.com/mailman/listinfo/freebase-discuss
>>>>>
>>>>
>>>>
>>>>
>>>> --
>>>> -Thad
>>>> http://www.freebase.com/view/en/thad_guidry
>>>>
>>>
>>>
>>>
>>> --
>>> -Thad
>>> http://www.freebase.com/view/en/thad_guidry
>>>
>>
>>
>>
>> --
>> -Thad
>> http://www.freebase.com/view/en/thad_guidry
>> _______________________________________________
>> You are receiving this message because you are subscribed to the Freebase-discuss mailing list.
>> To post a message to the list: Freebase-discuss at freebase.com
>> To unsubscribe, view archives, etc: http://lists.freebase.com/mailman/listinfo/freebase-discuss
>>
> _______________________________________________
> You are receiving this message because you are subscribed to the Freebase-discuss mailing list.
> To post a message to the list: Freebase-discuss at freebase.com
> To unsubscribe, view archives, etc: http://lists.freebase.com/mailman/listinfo/freebase-discuss
>