This project will consist of automating the process of DBpedia Neural Question Answering [1] model lifecycle management with the following tasks:
(1) make DBNQA models based on current dataset available to users;
(2) implement an evaluation service for DBpedia model evaluation with IRBench [2];
(3) integrate automatic build when changes occur in the DBpedia NSPM training dataset.
Description
This project will consist of automating the process of DBpedia Neural Question Answering [1] model lifecycle management with the following tasks:
(1) make DBNQA models based on current dataset available to users;
(2) implement an evaluation service for DBpedia model evaluation with IRBench [2];
(3) integrate automatic build when changes occur in the DBpedia NSPM training dataset.
Hello, I just get to know GSOC and this organization, but I’m pretty interested in this project. Will it be too late if I start to work on proposal now?
I’m Yiwei Dai, a senior student from China. I study Computer Science at Nanjing University and got admitted to Cornell University to continue the study as a master student in 2020 Fall, but I choose to gap for a year to do something cool because of the coronavirus. I have several projects experience in NLP and am working on my graduation paper concerning Forgetting Algorithm on ALC Ontology. I’m doing the warm-up tasks and try to show my potential to finish this project. Hope to get your reply. Thank you so much.
Hi
My name is Aman Kansal. I am a third-year undergraduate student at Indian Institute of Technology Bombay, India. I have done a variety of NLP and development research projects at my institute and am really interested in this project. I’ve started with the warmup tasks and also the proposal. Do we have a proposal template for this?
Thanks for your help.
(Sorry for joining the party late. Hope I’m not too late!)
Thank you for your remind. I’ve read about the rules carefully. I’m still a student when the result of proposal acceptance is released, so the eligibility should be satisfied.
I had a doubt regarding the project - are we aiming at extending KBox or building a neural model management framework from scratch?
Also @emarx, can you please point to some resources you think would help in better understanding the project, especially for IRBench and some similar work/papers
Thanks for your help