מתודולוגיה לניתוח מסעות לקוח דיגיטליים

.הסמינר יועבר ע״י ליהי בר-אל, סטודנטית לתואר שני במחלקה להנדסת תעשייה, אוניברסיטת תל אביב
ניתן לצפות גם דרך הזום

Abstract:
The enormous amounts of event log data collected from websites and mobile applications in recent years allow us to learn more about user behaviour patterns in digital systems.

In this work, we propose a methodology that aims to analyse thousands, if not millions, of digital event logs to derive actionable insights about user journeys. To this end, we use the Latent Dirichlet Allocation (LDA) algorithm to identify the most relevant types of user journeys. This is done by generating "topics" from sequences of events (sessions) that are transformed into clusters of sessions. We then capture sequential behavioural patterns among these clusters by representing them as Markov models. We present possible applications, such as finding the most representative paths and detecting actions that with high probability lead to undesirable outcomes.

We illustrate the implementation of the proposed methodology over a real-world dataset. The data contains event logs from a mobile application of a large HMO, that has more than 120K daily active users, generating ~5.5M events per day.

Bio:

Lihi Bar-El is an M.Sc student in the Department of Industrial Engineering in Tel Aviv University, under the supervision of Prof. Irad Ben-Gal. She holds a B.Sc in Industrial Engineering from Tel Aviv University. She works as a data scientist in a startup developing an AI powered LTV predictive platform