thanks Heiko! I will talk to you guys after reading these paper.
> 在 Feb 27, 2014，8:26 PM，Heiko Paulheim <[email protected]> 写道： > > Hi Zhang, > > thanks for your interest in the proposal! Here's a paper describing what's > already been done in DBpedia at the moment: > http://www.heikopaulheim.com/documents/iswc2013.pdf > > As you can see, there are some very basic mechanisms of type completion (only > based on links), which work surprisingly well. > > We have also experimented with Association Rules, with rather mixed results, > as shown in this prototype: > http://challenge.semanticweb.org/2012/submissions/swc2012_submission_15.pdf > > Any work extended these approaches (and reaching a good accuracy) is > appreciated. Ideas include, but aren't limited to > * including additional features (categories, other types, ...) > * including features from linked resoures (e.g., YAGO, Freebase) > * treating type completion as a hierarchical multi-class learning problem > * ... > > Hope that inspires you for some ideas. Feel free to come back with questions > and suggestions. > > All the best, > Heiko > > > > Am 27.02.2014 12:38, schrieb Marco Fossati: >> Hi Zhang and welcome on board! >> >> @Heiko, could you please provide some references to expand the idea proposal? >> >> Cheers! >> >>> On 2/26/14, 2:15 PM, 张天雷 wrote: >>> hi, everyone. >>> >>> This is Zhang Tianlei, a PHD candidate of Computer Science, Tsinghua >>> University @BeiJing, China. >>> >>> I spend most of my time building up knowledge bases, I've been worked >>> with MSRA EntityCube (StatSnowball) and Q20 (crowdsourcing) team for the >>> last two years and I use DBpedia, Freebase quite a lot during my daily >>> work, you guys have done a great work, and I really want to get involve >>> into this project, because I want to learn more, and I want more people >>> to know dbpedia, get benefits from it just like me. >>> >>> I am familiar with knowledge base completion techniques like pattern >>> based extraction, frequent item mining, and Distance Supervision. I use >>> a lot of statistical staff, such as Markov Logic Networks and CRF. And >>> I know needed NLP skills like POS-tag, NER, EntityLinking etc. >>> >>> I love python and java, I write most of my personal code in python like >>> twitter crawler and classfier, using gensim or scikit-learn, I even >>> teach these tools to other researchers. And for java, I write Android >>> codes and I use Stanford-NLP a lot. >>> >>> Personally I think axioms and common sense knowledge is very important >>> in knowledge mining, yet its remain a unexplored topic, and I hope that >>> more information about this topic could be forward to me, I want to get >>> involve with great projects. >>> >>> Thanks! >>> >>> >>> >>> ------------------------------------------------------------------------------ >>> >>> Flow-based real-time traffic analytics software. Cisco certified tool. >>> Monitor traffic, SLAs, QoS, Medianet, WAAS etc. with NetFlow Analyzer >>> Customize your own dashboards, set traffic alerts and generate reports. >>> Network behavioral analysis & security monitoring. All-in-one tool. >>> http://pubads.g.doubleclick.net/gampad/clk?id=126839071&iu=/4140/ostg.clktrk >>> >>> >>> >>> >>> _______________________________________________ >>> Dbpedia-gsoc mailing list >>> [email protected] >>> https://lists.sourceforge.net/lists/listinfo/dbpedia-gsoc > > -- > Dr. Heiko Paulheim > Research Group Data and Web Science > University of Mannheim > Phone: +49 621 181 2646 > B6, 26, Room C1.08 > D-68159 Mannheim > > Mail: [email protected] > Web: www.heikopaulheim.com > ------------------------------------------------------------------------------ Flow-based real-time traffic analytics software. Cisco certified tool. Monitor traffic, SLAs, QoS, Medianet, WAAS etc. with NetFlow Analyzer Customize your own dashboards, set traffic alerts and generate reports. Network behavioral analysis & security monitoring. All-in-one tool. http://pubads.g.doubleclick.net/gampad/clk?id=126839071&iu=/4140/ostg.clktrk _______________________________________________ Dbpedia-gsoc mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/dbpedia-gsoc
