[okfn-be] Request use cases - feedback - tool in development

Pieter Colpaert pieter at openknowledge.be
Mon Feb 19 11:00:59 UTC 2018


Hi Maxime,

For connecting the data package community with the Linked Data 
community, there happened some discussion recently on the Datahub.io 
gitter: https://gitter.im/datahubio/chat

Looks interesting in each case!

Kind regards,

Pieter


On 19-02-18 11:35, Maxime Gobert wrote:
>
> Hello Pieter,
>
> Thank you for your email.
>
> I did know the work if this lab, they are doing great work (I even use 
> their Open Data Portal Watch api as a source for my datasets 
> extraction feature)
>
> However I did not know about the odgraph tool yet. Looks promising, 
> thanks.
>
> I indeed thought about the problem of integrating linked data to the 
> Data Package format, it seems that the two are not fully compatible 
> yet, meaning that their tabular format does not support natively a URI 
> resource in addition of the value.
>
> This is still an open problem, I thought about maybe using the JSON-LD 
> format.
>
> My data package integration for now would be used to easily extract a 
> set of similar datasets found by the tool. This in order to eventually 
> be fed to another tool that specializes in something else like 
> visualization of similarities.
>
> Kind regards,
>
> Maxime
>
>
>
> Le 15-02-18 à 17:34, Pieter Colpaert a écrit :
>>
>> Hi Maxime!
>>
>> Sounds interesting! Some related work to that is from university of 
>> Vienna: http://data.wu.ac.at/odgraph/ → They extract geo-semantic 
>> labels from Open Datasets and create a search interface across this.
>>
>> Next, do you convert your dataset to Linked Data and then export it 
>> to the data package format? Because the latter really applies to non 
>> Linked Data. For what exactly is this useful?
>>
>> Finally, is there some code we can look into, or some dataset that 
>> was the end result of your process?
>>
>> Looking forward to meeting you at Open Belgium 2018!
>>
>> Pieter
>>
>>
>> On 15-02-18 11:48, Maxime Gobert wrote:
>>>
>>> Hello everyone,
>>>
>>> I'm a researcher at the University of Namur and I'm currently 
>>> working on a project on Data Interoperability and the link between 
>>> model and data, and I'm new to the Open Data subject.
>>>
>>> Regarding this subject I quickly oriented my work towards the Open 
>>> Data and Linked Data.
>>>
>>> How could the Open Data currently available be enriched with 
>>> semantic data (with the 5-star model of Tim Berners Lee in mind) ? 
>>> And How could a user of Open Data easily find datasets about a 
>>> particular subject (regarding the fact that there is a plethora of 
>>> data portals with limited search functionalities) ?
>>>
>>> Having those questions in mind I started developing a tool.
>>>
>>> The main functionalities of this tool are :
>>>
>>> -    Extraction of csv data sets via API of CKAN data portals.
>>>
>>> -    Analysis of column data types.
>>>
>>> -    Partial Semantic enrichment (currently with DBPedia resources)
>>>
>>> -    Search of analyzed files (via a text or a file). The tool will 
>>> find similar files (based on structure of the file, extracted meta 
>>> data, values, + other criteria possible not yet developed)
>>>
>>> -    Merger / Enhancer. The tool can currently merge (similar as 
>>> external sql join) two files based on two chosen columns. Future 
>>> goal is also to add columns to a file and get the data from the 
>>> identified dbpedia resources.
>>>
>>> -    Exporter. Finally the tool can export the result in the Data 
>>> Package Format of Frictionless data (https://frictionlessdata.io/)
>>>
>>> This is still a prototype and the reason I'm writing to you today is 
>>> that I'm searching for concrete use cases or users that could be 
>>> interested in such a tool in order to validate it's usefulness and 
>>> further develop it with new ideas and feedback.
>>>
>>> One example would be a startup that used or want to use Open Data, 
>>> what are their problems regarding the searching of Open Data? What 
>>> if they had such a tool?
>>>
>>> Anyway I could be interested in having contacts and discussions with 
>>> interested parties and see where it goes :)
>>>
>>> PS : I will be attending the Open Belgium event on 12 march.
>>>
>>> Thank you in advance for any feedback,
>>>
>>> Best regards,
>>>
>>> Maxime Gobert
>>>
>>>
>>> -- 
>>>
>>> Maxime GOBERT
>>> Chercheur
>>> PReCISE
>>>
>>> maxime.gobert at unamur.be <mailto:maxime.gobert at unamur.be>
>>> https://researchportal.unamur.be/fr/persons/maxime-gobert
>>>
>>> Université de Namur ASBL
>>> Rue de Bruxelles 61 - 5000 Namur
>>> Belgique
>>>
>>> Let’s respect the environment together.
>>> Only print this message if necessary!
>>>
>>>
>>>
>>> _______________________________________________
>>> okfn-be mailing list
>>> okfn-be at lists.okfn.org
>>> https://lists.okfn.org/mailman/listinfo/okfn-be
>>> Unsubscribe:https://lists.okfn.org/mailman/options/okfn-be
>>
>> -- 
>> +32486747122
>

-- 
+32486747122

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.okfn.org/pipermail/okfn-be/attachments/20180219/50358d0d/attachment-0003.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: domcngaeflejiodg.png
Type: image/png
Size: 9356 bytes
Desc: not available
URL: <http://lists.okfn.org/pipermail/okfn-be/attachments/20180219/50358d0d/attachment-0003.png>


More information about the okfn-be mailing list