About

Yoram has been working at DDCM as a doctoral student since 2018, after obtaining his degree as Master of Science in Computer Science Engineering summa cum laude earlier that year. Since then, he has been employed at DDCM as a doctoral student.

Current projects

Yoram’s research activities focus on data and information quality. Two parallel research projects that he has been working on can be distinguished to this end:

Cost-based analysis of data quality

Since the importance of high quality data is ever increasing, adequate tools for the analysis of data quality are needed. However, existing quality measures are typically not capable of quantifying differences between the quality of datasets. Moreover, they often approach data quality as a purely intrinsic property, thereby neglecting that the context in which data are used should be taken into account as well.

In this project, alternative and operational ways for quality analysis are proposed. This is done by analyzing the difficulties that are associated with solving tasks using the data. Yoram’s research activities hereby focus on setting up experimental designs for cost-based quality analysis. Apart from obtaining insights concerning the usefulness of a cost-based quality approach, conducting such experiments also provide a lot of information concerning data quality itself.

Monitoring of online news accuracy

A second project in which Yoram is involved focuses on a more practical setting in which data quality is relevant. In the past decades, news consumption has moved from reading traditional, paper-based newspapers to mainly online news consumption. This transition has been associated with a rise in misinformation and worrying trends such as fake news. In this project, tools are developed for the automated monitoring of the accuracy of professional online newspapers. This is done by monitoring edits made to news articles. A news tracker platform was developed where manual input of users concerning the purposes of edits made to articles was collected. The main idea of this platform is to use machine learning techniques to develop models for automated categorization of edits. Such models can be used to obtain estimates of the amount of errors corrected during certain periods of time or related to certain events. The models are publicly available at GitHub.

Previous projects

Yoram was also involved in the CryptoDrug project. Here, we composed a temporal database containing transactions made on several cryptomarkets.