Measurement of Users’ Experience on Online Platforms from Their Behavior Logs

摘要: Explicit measurement of experience, as mostly practiced, takes the form satisfaction scores obtained by asking questions to users. Obtaining response from every user is not feasible, responses are conditioned on questions, and provide only a snapshot, while experience journey. Instead, we measure values users’ click actions (events), thereby measuring for event. The without-asking-questions, combining recurrent neural network (RNN) with value elicitation event-sequence. platform environment modeled using an RNN, recognizing that user’s sequence has temporal dependence structure. We then elicit latent construct in this environment. offer two methods: one based rules crafted consumer behavior theories, another data-driven approach fixed point iteration, similar used model-based reinforcement learning. Evaluation comparison baseline show themselves good basis predicting conversion behavior, without feature engineering.