Hi Kamila, I'm probably late to the discussion because you have already made a nice progress on the proposal but I just wanted to put another topic on the table which in my opinion could be quite suitable for a GSoC project. It has to do with a new EntityHub Yard implementation. ATM there are (as far as I know) three different Yard implementations based on Solr, Clerezza and Sesame. First one is suitable to be used for Entity Linking and the others could be eventually used as classic TripleStore enabling for example SPARQL querying. We are still missing of a Yard implementation that could fulfil both uses cases with a single technology, basically a Triple Store with full text search capabilities. There are at least a couple of them available like Jena TDB + Jena Text (formerly Jena LARQ) or Stardog which is a commercial triple store providing a Community version as well.
I have been interested in this topic for years but never had the proper time to work on it. Apart from the Yard implementation, a new Entity Searcher would be to need implemented as well. @Rupert is a better authorised voice for this anyone as main developer of that part. My 2 cents, Rafa On Tue, Mar 27, 2018 at 5:55 AM Kamila Molina Orellana < kamila.molin...@gmail.com> wrote: > Hi Rupert, > > Thanks for your answer. > > I was seeing STANBOL-320 mostly as data cleaning/management issue. I will > try to describe the scope and possible solutions. > > However, until we define that, I think it is better to leave my proposal > with an open issue. Just to reiterate, I can say that I will exchange the > current SPARQL editor + an issue TBD (the deadline for the proposals is in > 12 hours). So, we can decide an issue or some work in the remaining time. I > said that because you can help me to estimate times, so I don't take an > issue that is too small or too big. > > Finally, this is my proposal [1] in case you have some comments. > > Regards, > ~Kamila. > > [1] > > https://docs.google.com/document/d/1WmropX2Bu_0g10VX3ZRE8Jil5kGb8N9RwhSCTs4KFww/edit?usp=sharing > > On Mon, Mar 26, 2018 at 5:32 AM, Rupert Westenthaler < > rupert.westentha...@gmail.com> wrote: > > > Hi, > > > > Disambiguation is interesting for sure, but also a very broad topic. > > So please make sure to describe scope and possible solutions (e.g. > > approach + frameworks) to the problem. > > > > A dummy example based on STANBOL-320 this could be > > > > * approach: process extracted named entities; filter all with more as > > 5 words and 50 chars > > * frameworks: none > > > > A more realistic example could be > > > > * approach: summary of the approach + links to some papers this > > approach is based on or related to > > * frameworks: links to the frameworks used to implement the approach. > > If required also links to the datasets needed for learning ... > > > > hope this helps > > best > > Rupert > > > > On Sun, Mar 25, 2018 at 11:51 PM, Kamila Molina Orellana > > <kamila.molin...@gmail.com> wrote: > > > Hi, > > > > > > I am working into my proposal. I think it would be a good a idea to > > define > > > some ground work that I can work after moving to Squebi. I was looking > at > > > some of the issues, and there are some really interesting However, I am > > > afraid mostly for the time. Specifically, since I have been doing and > > > learning about semantic web and linked data at my school, I was > > interested > > > in the issue of entity disambiguation [1]; it looks really exciting and > > > challenging. Beyond, that I can learn more about Stanbol, I can apply > > some > > > interesting concepts. > > > > > > Or maybe this one about filtering some results of NED engine [2]. Do > you > > > have any other suggestions? Do think It would be to better to fix some > > > others issues that are more critical for the project? > > > > > > Finally, I found this mentoring guide from Apache for GSoC in case > > Antonio > > > or Rafa would like to help Andreas to co-mentoring in GSoC. > > > > > > Hope to hear back from you soon. > > > > > > Regards, > > > ~Kamila. > > > > > > > > > [1] https://issues.apache.org/jira/browse/STANBOL-1183 > > > [2] https://issues.apache.org/jira/browse/STANBOL-320 > > > [3] https://community.apache.org/guide-to-being-a-mentor.html > > > > > > On Mon, Mar 12, 2018 at 9:58 PM, Kamila Molina Orellana < > > > kamila.molin...@gmail.com> wrote: > > > > > >> Hi, > > >> > > >> Great Antonio , thanks!! :D Then, I would keep working in my proposal > > and > > >> share with you for some feedback. Also, can we define some issues > that I > > >> can work on? Or what could be the extra work? Until then, I will look > at > > >> some issues/code to get some idea of the work. > > >> > > >> Regards, > > >> ~Kamila. > > >> > > >> On Mon, Mar 12, 2018 at 5:11 AM, Antonio David Pérez Morales < > > >> adperezmora...@gmail.com> wrote: > > >> > > >>> Hi Andreas, > > >>> > > >>> I don't have a plenty of time this year, but I think I can find a bit > > of > > >>> time and act as a mentor for this GSoC year. > > >>> > > >>> Regards > > >>> > > >>> 2018-03-10 4:53 GMT+01:00 Kamila Molina Orellana < > > >>> kamila.molin...@gmail.com> > > >>> : > > >>> > > >>> > Hi all, > > >>> > > > >>> > Rafa the idea would be use Squebi as the default Sparql Endpoint > > >>> instead of > > >>> > the one being used now [1]. However, Rupert point out that task > > would be > > >>> > really short. So, I say we can add some issues too. Or what do you > > >>> suggest? > > >>> > > > >>> > [1] https://issues.apache.org/jira/browse/STANBOL-1327 > > >>> > > > >>> > Regards, > > >>> > ~Kamila. > > >>> > > > >>> > On Fri, Mar 9, 2018 at 9:49 AM, Rafa Haro <rh...@apache.org> > wrote: > > >>> > > > >>> > > Hi Andreas, > > >>> > > > > >>> > > Can you or the student elaborate here a little bit the idea for > the > > >>> > > project? > > >>> > > > > >>> > > Thanks > > >>> > > > > >>> > > On Fri, Mar 9, 2018 at 8:26 AM Andreas Kuckartz < > > a.kucka...@ping.de> > > >>> > > wrote: > > >>> > > > > >>> > > > Hi Kamila and Rupert, > > >>> > > > > > >>> > > > Sorry for my very belated reply. > > >>> > > > > > >>> > > > The intention of tagging the issue with GSOC 2018 was that it > > could > > >>> be > > >>> > > > resolved in a GSOC-project. > > >>> > > > > > >>> > > > Yes, it alone would not be enough for such a project and > > therefore > > >>> > > > significantly more work would need to be included. > > >>> > > > > > >>> > > > The main problem seems to be to have two mentors. > > >>> > > > > > >>> > > > Is there any of the other Stanbol developers who can and likes > to > > >>> fill > > >>> > > > such a role? It would really help to support Stanbol. > > >>> > > > > > >>> > > > Only then should we attempt to develop the scope of the GSOC > > >>> project. > > >>> > > > > > >>> > > > Cheers, > > >>> > > > Andreas > > >>> > > > --- > > >>> > > > > > >>> > > > Kamila Molina Orellana schrieb: > > >>> > > > > Thanks Rupert for you answer. I understand it. I was really > > >>> > interested > > >>> > > in > > >>> > > > > Stanbol, so I guess I will keep looking at some other project > > for > > >>> > GSoC. > > >>> > > > > However, If anyone is interested in mentoring, I am still > > looking > > >>> > > forward > > >>> > > > > to contribute. > > >>> > > > > > > >>> > > > > Regards, > > >>> > > > > ~Kamila. > > >>> > > > > > > >>> > > > > On Mon, Mar 5, 2018 at 7:02 AM, Rupert Westenthaler < > > >>> > > > > rupert.westentha...@gmail.com> wrote: > > >>> > > > > > > >>> > > > >> Hi Kamila > > >>> > > > >> > > >>> > > > >> Stanbol has very little ongoing development - mostly fixing > > bugs. > > >>> > > > >> While their are for sure topics of interests - especially > > >>> related to > > >>> > > > >> information extraction and classification - I do not know if > > we > > >>> > would > > >>> > > > >> be able to find 2 Developers plan to be active enough for > > >>> Mentoring. > > >>> > > > >> > > >>> > > > >> For me the GSoC 2018 timeline is troublesome as I will be > > mostly > > >>> > > > >> offline in final evaluation period (6 - 21 August). So I > would > > >>> have > > >>> > a > > >>> > > > >> hard time providing list minute feedback and would have a > hard > > >>> time > > >>> > to > > >>> > > > >> do the final evaluation. > > >>> > > > >> > > >>> > > > >> best > > >>> > > > >> Rupert > > >>> > > > >> > > >>> > > > >> > > >>> > > > >> On Sun, Mar 4, 2018 at 6:31 PM, Kamila Molina Orellana > > >>> > > > >> <kamila.molin...@gmail.com> wrote: > > >>> > > > >>> Hi Rupert. > > >>> > > > >>> > > >>> > > > >>> Thanks so much Rupert for your answer. Do you think the > > project > > >>> > will > > >>> > > go > > >>> > > > >> for > > >>> > > > >>> GSoC? I have been looking at Squebi, and if it is too > short, > > we > > >>> can > > >>> > > try > > >>> > > > >> to > > >>> > > > >>> fix some other issue. > > >>> > > > >>> > > >>> > > > >>> Regards, > > >>> > > > >>> ~Kamila. > > >>> > > > >>> > > >>> > > > >>> On Wed, Feb 28, 2018 at 1:09 AM, Rupert Westenthaler < > > >>> > > > >>> rupert.westentha...@gmail.com> wrote: > > >>> > > > >>> > > >>> > > > >>>> Hi Kamila, > > >>> > > > >>>> > > >>> > > > >>>> Sorry for the late response, but I wanted wait for Andreas > > to > > >>> > answer > > >>> > > > >>>> this as I do not really agree with him tagging > STANBOL-1327 > > >>> with > > >>> > > > >>>> GSoC2018. > > >>> > > > >>>> > > >>> > > > >>>> IMHO this would be not a good topic as it is much to > simple > > and > > >>> > > small > > >>> > > > >>>> (I would expect a maximum of 2 working days to complete > this > > >>> > > feature). > > >>> > > > >>>> Maybe Andreas can add some comments about his intentions > > when > > >>> > > marking > > >>> > > > >>>> this issue with GSoC2018. > > >>> > > > >>>> > > >>> > > > >>>> best > > >>> > > > >>>> Rupert > > >>> > > > >>>> > > >>> > > > >>>> On Mon, Feb 2 > <https://maps.google.com/?q=%3E+%3E%3E%3E%3E+On+Mon,+Feb+2&entry=gmail&source=g>6, > 2018 at 4:58 AM, Kamila Molina Orellana > > >>> > > > >>>> <kamila.molin...@gmail.com> wrote: > > >>> > > > >>>>> Hi, > > >>> > > > >>>>> > > >>> > > > >>>>> I have been working in my proposal and wanted to make the > > >>> most of > > >>> > > > >>>> bounding > > >>> > > > >>>>> period by interacting with th > <https://maps.google.com/?q=interacting+with+th&entry=gmail&source=g>e > community and the tool. I > > have > > >>> > been > > >>> > > > >>>> working > > >>> > > > >>>>> in the proposal and would like to share it with my > > prospective > > >>> > > > mentor. > > >>> > > > >>>>> Anyone from Apache who would like to mentorship? > > >>> > > > >>>>> > > >>> > > > >>>>> Regards, > > >>> > > > >>>>> ~Kamila. > > >>> > > > >>>>> > > >>> > > > >>>>> On Tue, Feb 20, 2018 at 9:31 PM, Kamila Molina Orellana < > > >>> > > > >>>>> kamila.molin...@gmail.com> wrote: > > >>> > > > >>>>> > > >>> > > > >>>>>> Hi all, > > >>> > > > >>>>>> > > >>> > > > >>>>>> I am interested in participating in GSOC 2018 and been > > >>> looking > > >>> > at > > >>> > > > the > > >>> > > > >>>>>> issue and playing a bit with Stanbol. I am a student > from > > the > > >>> > > > >>>> University of > > >>> > > > >>>>>> Cuenca. I have been working with Semantic technologies > and > > >>> > > currently > > >>> > > > >> in > > >>> > > > >>>> my > > >>> > > > >>>>>> third year of college. > > >>> > > > >>>>>> > > >>> > > > >>>>>> Well just a quick overview of what I understand, please > > >>> correct > > >>> > me > > >>> > > > >> if I > > >>> > > > >>>> am > > >>> > > > >>>>>> wrong. The idea will be to exchange the actual Sparql > > >>> Endpoint > > >>> > for > > >>> > > > >>>> Squebi. > > >>> > > > >>>>>> Then when you go to /sparql, you'll have Squebi > > >>> functionalities, > > >>> > > > >> right? > > >>> > > > >>>>>> > > >>> > > > >>>>>> I see you use Fremaker to build the website, OSGI to > load > > >>> > modules. > > >>> > > > >> The > > >>> > > > >>>>>> actual Sparql Endpoint is loading in bundle/list.xml ( > > >>> > > > >>>>>> org.apache.stanbol.commons.web.sparql), but when is this > > >>> file > > >>> > > > >> loaded? > > >>> > > > >>>>>> Where is the module that manages the Sparql Services? I > > mean, > > >>> > the > > >>> > > WS > > >>> > > > >> to > > >>> > > > >>>>>> make updates and select queries. > > >>> > > > >>>>>> > > >>> > > > >>>>>> > > >>> > > > >>>>>> Regards, > > >>> > > > >>>>>> ~Kamila. > > >>> > > > >>>>>> > > >>> > > > >>>> > > >>> > > > >>>> > > >>> > > > >>>> > > >>> > > > >>>> -- > > >>> > > > >>>> | Rupert Westenthaler > > >>> rupert.westentha...@gmail.com > > >>> > > > >>>> | Bodenlehenstraße 11 > > >>> > > ++43-699-11108907 <+43%20699%2011108907> > > >>> > > > <+43%20699%2011108907> > > >>> > > > >>>> | A-5500 Bischofshofen > > >>> > > > >>>> | REDLINK.CO .............................. > > >>> > > > >> .............................. > > >>> > > > >>>> .............. > > >>> > > > >>>> | http://redlink.co/ > > >>> > > > >>>> > > >>> > > > >> > > >>> > > > >> > > >>> > > > >> > > >>> > > > >> -- > > >>> > > > >> | Rupert Westenthaler > > rupert.westentha...@gmail.com > > >>> > > > >> | Bodenlehenstraße 11 > > >>> > ++43-699-11108907 <+43%20699%2011108907> > > >>> > > > <+43%20699%2011108907> > > >>> > > > >> | A-5500 Bischofshofen > > >>> > > > >> | REDLINK.CO > > >>> > > > ............................................................ > > >>> > > > >> .............. > > >>> > > > >> | http://redlink.co/ > > >>> > > > >> > > >>> > > > > > > >>> > > > > > >>> > > > > > >>> > > > > >>> > > > >>> > > >> > > >> > > > > > > > > -- > > | Rupert Westenthaler rupert.westentha...@gmail.com > > | Bodenlehenstraße 11 ++43-699-11108907 > <+43%20699%2011108907> > > | A-5500 Bischofshofen > > | REDLINK.CO > ............................................................ > > .............. > > | http://redlink.co/ > > >
