https://github.com/smartdataanalytics/qacomponents-analysis
https://github.com/smartdataanalytics/qacomponents-analysis
Last synced: about 2 months ago
JSON representation
- Host: GitHub
- URL: https://github.com/smartdataanalytics/qacomponents-analysis
- Owner: SmartDataAnalytics
- Created: 2018-06-19T14:20:50.000Z (almost 7 years ago)
- Default Branch: master
- Last Pushed: 2020-06-28T13:34:54.000Z (almost 5 years ago)
- Last Synced: 2025-01-22T06:22:43.418Z (3 months ago)
- Language: R
- Size: 1.82 MB
- Stars: 0
- Watchers: 17
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# QAComponents-analysis
This is the Github repository for the submitted JWS paper "No One is Perfect: Analysing the Performance of Question Answering Components over the DBpedia Knowledge Graph".
Following are the folders in the repository for various experiments in the paper: The original Frankenstein repository: [Frankenstein Framework](https://github.com/WDAqua/Frankenstein).
In Frankenstein repository, you will be pointed to the Qanary components, which are part of this analysis paper. You can directly find the components in [Qanary Components](https://github.com/WDAqua/Qanary-question-answering-components). All the components are recently updated to ensure reproducibility.
One can educate them in detail about Qanary methodology and its usage on: [Qanary Methodology](https://github.com/WDAqua/Qanary). It also includes how to add a new component and run a pipeline.
For non-expert users, a majority of components are available to access via a web UI: [Frankenstein Web UI](http://frankenstein.qanary-qa.com/).
The *RawData* folder contains all raw data produced in our experiments that have been used in the analysis. In particular, the raw data is organized in the following folders:
- *Component Performance*: This folder contains all the files of the different question Answering Tasks (QA Tasks). F-score for each question per component
is given in the files.
- *Component Performance Aggregated*: This folder sums up the f-score values of the component per question feature for various QA tasks.
- *Top Count*: This folder contains values of the number of questions answered by Top 1, top 2...top n components together for each task.
- *Unanswered*: This folder contains all the unanswered questions, and the associated question features.
Finally, the folder *Rplots* contains the R scripts used to generate the plots of the paper, as well as the generated plots in pdf format.