WeekThree | Literature Survey

It’s the third week now, I have gone through a lions part of the existing codebase that I am going to work on for the next few months.

While doing my older project, I learnt an important aspect of working on something new and that was to understand the bleeding edge of technology that pertains to your area of work. While creating the proposal I did go through a lot of papers but while solely working on the project I decided to dig a bit deeper.

Work done this week:


Literature Survey.

The first stop was to re-read the papers I already read that were part of this project:

These papers forma the very base of this project idea, and it was important to understand them properly before diving into the code.


Now there were 2 important aspect to the final goal of this project that I had decided to focus on:

While searching for complex Q&A I found about the LCQuAD dataset:

LCQuAD

{
        "_id": "1701", 
        "corrected_question": "Which architect of Marine Corps Air Station Kaneohe Bay was also tenant of New Sanno hotel /'", 
        "intermediary_question": "What is the <architect> of the <Marine Corps Air Station Kaneohe Bay> and <tenant> of the <New Sanno Hotel>", 
        "sparql_query": " SELECT DISTINCT ?uri WHERE { <http://dbpedia.org/resource/Marine_Corps_Air_Station_Kaneohe_Bay> <http://dbpedia.org/property/architect> ?uri. <http://dbpedia.org/resource/New_Sanno_Hotel> <http://dbpedia.org/ontology/tenant> ?uri} ", 
        "sparql_template_id": 16
    }

QALD Benchmark

QALD is a series of evaluation campaigns on question answering over linked data. So far, it has been organized as an ESWC workshop and an ISWC workshop as well as a part of the Question Answering lab at CLEF.

Checking out the previous competitors in QALD dataset:

WDAqua-core1: A Question Answering service for RDF Knowledge Bases, brief workflow is given below:


There were other approaches too, but most of them were rule based in nature.

Conclusion

Index Page