Re: GSoC Final Project Deadlines

On Sep 4, 2013, at 2:16 PM, vikash agrawal <vikashagrawal1990@gmail.com> wrote:

> 
> 
> Thanks for bringing this up.
> 
> On Thu, Sep 5, 2013 at 1:50 AM, Gregg Kellogg <gregg@greggkellogg.com> wrote:
>> It's reasonable, but pretty shallow, and certainly doesn't follow the best practices of Linked Data. For example, it shows that Wikia is one of companies I work for, but just has a blank node with name "Wikia". Looking at the returned JSON from LinkedIn, though, there's enough information to create an actual IRI for it. Looking at the returned JSON, I see the following for positions:
>> 
>>    "positions": {
>>       "_total": 15,
>>       "values": [{
>>         "company": {
>>           "id": 157252,
>>           "industry": "Internet",
>>           "name": "Wikia",
>>           "size": "51-200 employees",
>>           "type": "Privately Held"
>>         },
>>         "id": 291520813,
>>         "isCurrent": true,
>>         "startDate": {
>>           "month": 6,
>>           "year": 2012
>>         },
>>         "summary": "Adding structure to wikis.",
>>         "title": "Linked Data Consultant, Advisory Board member"
>>     ...
>> 
>> This is enough to link to the company, and provide some context for the work I've done with them. At the least, I would create an @id referencing the Wikia profile. This would just be <http://www.linkedin.com/company/157252>, which you can figure out given that it is a "company" position, and has an id. Same would hold for other references, such as education.
> 
> Yes, initially I did thought of doing this by generating IRI's as you said -> linkedin.com/company/id for companies and linkedin.com/edu/school?id= but it seems it will also generate misleading url incase where the organisations (company, school..) dont have their profile on LinkedIn. And we tend to get [1],[2] and [3]. A 404 page saying Company/Scholl does not exist. So as a result I thought it would be good to have it this way instead of having wrong information.

AFAIK, this is a valid URL, it just hasn't been set up. As
I recall, it does give you a template page. In any case, I suppose you could check the URL first. It certainly isn't wrong to use it.

Gregg

>> 
>> Also, the "owns" information isn't too useful. It like your setting a type of <http://schema/url>, which is a property, not a type. And your using the account URL as the name and throwing the name away. For example, the following:
>> 
>> {
>>   "@type": "url",
>>   "name": "http://www.kellogg-assoc.com"
>> }
> 
> Thanks for the notification
>  
>> Where the information from LinkedIn is:
>> 
>>     "memberUrlResources": {
>>       "_total": 3,
>>       "values": [{
>>         "name": "Personal Website",
>>         "url": "http://www.kellogg-assoc.com"
>>       }, {
>>         "name": "Laudits",
>>         "url": "http://www.laudits.com/pub/1162/gregg-kellogg/7trdv"
>>       }, {
>>         "name": "Twyla",
>>         "url": "http://www.twylah.com/Gkellogg?utm_source=New+users+week+of+20120319&utm_campaign=4d564b7e48-New_20120321&utm_medium=email"
>>       }]
>>     }
>> 
>> From that, I would think that you'd do more the following:
>> 
>> {
>>   "@id": "http://www.kellogg-assoc.com",
>>   "name"Personal Website"
>> }
> 
> Fixed: Yes, this is done. Can you please check it once to verify if this is what you intended.
> 
> 
>> You might be able to figure out that it has a @type of schema:OwnershipInfo, as that is the range of schema:owns (look at http://schema.org/owns).
> 
> My bad, I dint realise this. I have rectified it to OwnershipInfo.
> 
> So, did you encounter any more issues? And are things looking ok now?
> 
> Regards
> ~Vikash
> 
> [1] - https://dl.dropboxusercontent.com/u/5278881/GSoC/Companies404.png
> [2] - https://dl.dropboxusercontent.com/u/5278881/GSoC/School404.png
> [3] - https://dl.dropboxusercontent.com/u/5278881/GSoC/School404-1.png
> 
> 
> Regards
> ~Vikash 
>> Gregg Kellogg
>> gregg@greggkellogg.net