Re: Historical events

I could be wrong, but I think the distinction is that Wikidata and LCSH are
centralized stores of instance information, so if they want to say some of
their events are time periods they can base it on any definition of time
period they please. Schema contains no instance information, it's just a
controlled vocabulary for distributed data, so the terms need to have
agreed meaning. Does that sound right to people with more experience here?

Anthony

On Tue, Jun 19, 2018 at 9:23 AM Vicki Tardif <vtardif@google.com> wrote:

> You are correct that schema.org is a controlled vocabulary, which is
> precisely why I think we should avoid modifiers like 'significant'. It is
> up to authors to describe their data and readers to determine whether the
> data is relevant to their needs. For example an event that is significant
> on my calendar probably does not rise to the level of Wikidata notability,
> but both could in theory use schema.org to describe Events.
>
> When in doubt, I think it is fruitful to look at how other
> ontologies/controlled vocabularies are organized. Wikidata separates out
> time periods from events. As do LCSH and DDC.
>
> One may want to describe the collection of battles as a super-event called
> WWII. One may also find it useful to label a time period WWII. For example,
> if describing someone as being born in 1941, it is more useful to describe
> that era as a time period.
>
> - Vicki
>
>
> On Tue, Jun 19, 2018 at 12:02 PM Anthony Moretti <
> anthony.moretti@gmail.com> wrote:
>
>> I agree with the need for more specific types of events, I've been
>> arguing myself for Trip and ParcelDelivery to be moved to Event from
>> Intangible.
>>
>> Schema is a controlled vocabulary
>> <https://en.wikipedia.org/wiki/Controlled_vocabulary> though, so the
>> point is to have agreed meaning on terms. In the same vein I've been
>> arguing that CreativeWork, Product, and Intangible don't have agreed
>> meaning and therefore also don't belong in Schema. Terms like
>> "significant", "important", or "special" declare a subset of things, and to
>> give an example people do something similar when they star or flag emails.
>> I should be allowed to use stars, and you should be allowed to use stars,
>> but if we combined our mail we'd likely need to re-star things based on an
>> agreed meaning of stars, or alternatively not do it at all.
>>
>> On periods, eras, epochs, ages, etc., to me they seem the same as events
>> - portions of time, sometimes exact sometimes not, during which other
>> events happened or not, and with a name - just normally longer. "Mike's
>> birthday party", an event I'd argue according to most people, is nothing
>> but a portion of time during which other events happened - "Mike had a
>> shot", "Mike cut and served cake" etc. In addition to being an event (armed
>> conflict) would World War II also be considered a period or era? It's too
>> ambiguous in my view to go in a controlled vocabulary, hence the suggestion
>> for the additional duration categories.
>>
>> Anthony
>>
>> On Tue, Jun 19, 2018 at 7:08 AM Vicki Tardif <vtardif@google.com> wrote:
>>
>>> I don't think we should be the arbiter of 'significance' as we are
>>> writing the schema, not storing the data. Authors can apply the type as
>>> they see fit and it is up to readers to reconcile the data and determine if
>>> the event is significant for their purposes.
>>>
>>> With that said, I am not opposed to more specific types like
>>> ArmedConflict if that serves authors.
>>>
>>> I am opposed to adding things like 'EventLongerThan1Year'. These are
>>> time periods (or eras if we'd rather). We can add an 'approximateStartDate'
>>> and 'approximateEndDate' to allow for setting the length. Otherwise, the
>>> lengths are arbitrary and serve no purpose to put the time period in
>>> context.
>>>
>>> - Vicki
>>>
>>>
>>> On Mon, Jun 18, 2018 at 10:20 PM Anthony Moretti <
>>> anthony.moretti@gmail.com> wrote:
>>>
>>>> After looking at it more closely I think I agree with Martin, unless
>>>> there's general agreement on the meaning of a type it probably shouldn't be
>>>> included.
>>>>
>>>> "Significance" and "importance" are relative and only mean something if
>>>> there is agreement on several things: the group being compared, the
>>>> attribute being compared, and some qualifying percentile, e.g. top 5% of
>>>> mountains by height, top 10% of people by name recognition, top 1% of
>>>> websites by views.
>>>>
>>>> Particularly about events though, maybe duration is an interesting
>>>> attribute, quite long events becoming named periods of history. With the
>>>> following types as a partial example, you could look for archaeological
>>>> events longer than a certain duration, and the results might possibly be
>>>> common named periods of archaeology.
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>>
>>>> *Event    AnthropologicalEvent    ArchaeologicalEvent    ArmedConflict
>>>>   CosmologicalEvent    GeologicalEvent    PoliticalEvent    ReligiousEvent
>>>>   EventLongerThan1Year        EventLongerThan10Years
>>>>   EventLongerThan100Years                EventLongerThan1000Years
>>>>           EventLongerThan10000Years
>>>>   EventLongerThan100000Years
>>>>   EventLongerThan1000000Years
>>>>   EventLongerThan10000000Years
>>>>   EventLongerThan100000000Years
>>>>   EventLongerThan1000000000Years
>>>>   EventLongerThan10000000000Years*
>>>>
>>>>
>>>> (These duration categories cover the age of the universe.)
>>>>
>>>> Could this be a solution to classifying and finding "historically
>>>> significant" events?
>>>>
>>>> Anthony
>>>>
>>>> On Mon, Jun 18, 2018 at 4:22 PM <Simon.Cox@csiro.au> wrote:
>>>>
>>>>> Does it have a name? If so, it is ‘significant’. That’s about it
>>>>> really.
>>>>>
>>>>>
>>>>>
>>>>> *From:* Thad Guidry [mailto:thadguidry@gmail.com]
>>>>> *Sent:* Tuesday, 19 June, 2018 03:42
>>>>> *To:* Muri, Allison <allison.muri@usask.ca>
>>>>> *Cc:* Richard Wallis <richard.wallis@dataliberate.com>; Anthony
>>>>> Moretti <anthony.moretti@gmail.com>; Cox, Simon (L&W, Clayton)
>>>>> <Simon.Cox@csiro.au>; schema.org Mailing List <public-schemaorg@w3.org
>>>>> >
>>>>> *Subject:* Re: Historical events
>>>>>
>>>>>
>>>>>
>>>>> Allison,
>>>>>
>>>>>
>>>>>
>>>>> Does it help to think this way ?
>>>>>
>>>>>
>>>>>
>>>>> *A Significant* - some Thing of importance to someone or some group.
>>>>>
>>>>> A Significant Other - some Person of importance to another Person.
>>>>>
>>>>>
>>>>>
>>>>> Historical*Significant* - some Thing that has *historical* importance
>>>>> to someone or some group.
>>>>>
>>>>>
>>>>>
>>>>> -Thad
>>>>>
>>>>>
>>>>>
>>>>