Software Resources Repository

Details/Description:

Project with services called (either directly or indirectly) by the Knowledge Integration component and functions which run in predefined time intervals. Services include the social media retrieval, the topic detection, the passage retrieval for health-information, the text/document retrieval for newspapers and diabetic recipes, the named entities and concept extraction, the relation extraction and the KRISTINA-tailored location-based search services (i.e., events, nearest places of interest and weather forecast service). Functions run through this application are the German newspaper crawling, a special case of Web crawling that utilizes the available RSS feeds and the tweet classification module which runs the python scripts of the corresponding repository to update the Twitter crawling database.

Link:

https://github.com/MKLab-ITI/KRISTINA/tree/master/WebKIndex/

Details/Description:

Desktop application for the social media crawling, content scraping and indexing modules of the KRISTINA pipeline that are performed offline.

Link:

https://github.com/MKLab-ITI/KRISTINA/tree/master/KIndex/

Details/Description:

Python code for the tweet classification module in the frame of the KRISTINA project (http://kristina-project.eu/en/). It includes the code for both the experimental setup and the service itself which updates the category field of the recently crawled Twitter posts.

Link:

https://github.com/MKLab-ITI/KRISTINA/tree/master/CategoryClassification/

Details/Description:

Java code of the complete Semantic Framework. More specifically, the following modules are available as a single bundle: conversational awareness, question answering, fusion, knowledge integration, knowledge base (GraphDB triple store), as well as auxiliary services and utilities that implement the interaction with other components and modules of the overall KRISTINA framework.

Link:

https://github.com/MKLab-ITI/KRISTINA/tree/master/SemanticFramework/

Details/Description:

Scripts in the R programming language that implement the topic detection framework. They take as input a folder with text files and return the clustering result in json format. The difference between the two scripts is that the one handles texts in German language, while the other handles texts in Turkish.

Link:

https://github.com/MKLab-ITI/KRISTINA/tree/master/WebKIndex/topicDetectionFiles

Details/Description:

Code for calling the concept extraction module in the frame of the KRISTINA project (http://kristina-project.eu/en/).

Link:

https://github.com/MKLab-ITI/KRISTINA/tree/master/KIndex/KIndex/src/main/java/gr/iti/mklab/kindex/ConceptExtraction

Details/Description:

VERGE is a hybrid interactive video retrieval system, which is capable of searching into video content by integrating different search modules that employ visual- and textual-based techniques.

Link:

https://github.com/MKLab-ITI/verge