The ethical data-driven research lab

Maayan Zhitomirsky-Geffet received her PhD in natural language processing from the Hebrew University of Jerusalem (2006). In the past decade, she has been leading the field of ethical data-driven research in the Department of Information Science at Bar Ilan University. Dr. Zhitomirsky-Geffet has developed the ethical approach to representing, organizing and analyzing information based on multi-viewpoint ontologies. Her research deals with the development of ethical ontological models that serve as a tool for identifying and analyzing opinions and assessing biases in large textual corpora. The ethical approach may help identify various types of biases in the data (e.g. cultural, gender-oriented, religious, political, temporal, and more) in various domains (e.g., art, historical literature, politics, labor market, online personal branding, communication, academic promotion and influence, education) and balancing them by increasing the diversity of voices in knowledge representation and harmonization of different perspectives.

According to this approach, artificial intelligence algorithms will present the user with relevant information based on the "related yet different or complementary information" relevance criterion that allows for diversity, inclusion and balance rather than on the criterion of similarity to the previous content in which a user was engaged. That is, retrieval and personalization of information will be done on the basis of relevance defined as information that is “related yet different” in order to broaden the horizons of the user in the subject of his interest rather than narrowing his vision and views. This approach can neutralize the “rich getting richer” effect on the web that causes distortions and biases in the ranking of information importance by algorithms and ultimately also by humans who are exposed to these automatic rankings directly or indirectly. This ethical and inclusive approach to the organization and retrieval of information aims to prevent the spread and increase of the influence of hate content and fake news on the Internet, and thus also to prevent social polarization and even harm to democracy and civil war. Dr. Zhitomirsky-Geffet's laboratory also investigates other aspects of ethics in the use of technology, such as the application of multi-viewpoint ontological models for quantitative analysis (remote reading) of corpora in the humanities and social sciences, online user privacy and addiction to information technology.

Prof. Zhitomirsky-Geffet is a research partner in the European Time Machine Project, with the aim of building a multi-viewpoint ontology for Jerusalem and other multicultural cities, in the Jerusalem National Knowledge Center of the Ministry of Science, and in the European CSI-COP Project “citizen scientists investigating trackers on the web”. She also serves as a chair of the ethics committee of the Humanities and Jewish studies faculties in Bar-Ilan University.

Dr. Zhitomirsky-Geffet's research was presented at the top conferences such as ACL, WWW and ASIS&T, and her articles were published in various prestigious journals such as Computational Linguistics and JASIS&T.