In collaboration with Dr. Yoav Lehahn and Dr. Daniel Sher of the School of Marine Sciences. We are constructing a system for aggregation and sharing of marine research datasets in order to provide researchers the ability to fuse different datasets and receive combined dataset. Current research state aims to solve the problem by means of ontology-based data integration.
Some of the earliest and most important texts in human history were created in the middle east, often regarded as the cradle of civilization. Over recent years, document digitization methods have dramatically increased the number of digitally available historical texts in Semitic languages. Concurrently, a number of structured data sources have emerged in both Hebrew (e.g., Kima, DiJeSt ) and Arabic (e.g., al-Ṯurayyā Gazetteer). However, these sources remain disconnected and no unifying interface exists that allows either distributed search engines or human researchers to ask data-driven questions over multiple sources irrespective of their source language. The objectives of this project are to 1) Provide schema matching and multilingual entity resolution at scale to build a comprehensive multi-lingual knowledge graph from existing Semitic structured sources. 2) Enrich this knowledge graph with entities and relations extracted from Semitic texts (unstructured sources). 3) Provide a geo-temporal visualization and a query platform that humanities researchers can use to ask structured research questions. Project website: https://mehdie.org/
The Fantastic art (FA) artistic genre is, by far, the most popular genre among the younger crowd. Although FA is in high demand on the web, and in the gaming, literature, and tattooing industries, it suffers from a lack of academic attention. The “Fantastic Art Master” project will be the first time an entire artistic genre will receive an academic computerized examination to base upon its core research upon. It will offer an in-depth understanding of the artistic influences on the genre, of past and multicultural global material art, and form connections between museums and the popular new media art. In order to accumulate and curate a massive artistic corpus structured as a knowledge graph. A data organization structure that allows the integration of facts from multiple sources with an underlying ontology, enabling inferred knowledge to be generated from both the collected data and from its connections with other, related, knowledge graphs. The collected corpus will also serve future efforts to use machine learning to identify, classify and map the FA genre. This project is in collaboration with Dr. Sharon Khalifa-Gueta, Dr. Moshe Lavee, and Prof. Tsvi Kuflik.
https://www.sharon-dragon.com/
The Ontobuilder Research Environment (ORE ) is designed to provide students and researchers in schema matching and related fields a jump-start in designing and running experiments. At the data management lab, we develop ORE and use it to advance basic research in schema matching.
Do you know how your reflection looks like in social media? In collaboration with Dr. Reout Arbel of the Department of Education, we are developing a reflective digital model of a person's social and news streams (online reflection). We aim to evaluate how this reflection can be used to assess the sentiments a person receives from online sources on a continuous basis.