Search papers, labs, and topics across Lattice.
This paper introduces an automated method for generating diverse user personas for testing interview dialogue systems, addressing the labor-intensive nature of manual persona creation. By leveraging a large language model, the authors assign distinct personality traits to these personas, enhancing the variability of simulated user behaviors. Experimental results demonstrate that the generated personas lead to a richer diversity in utterances, improving the effectiveness of dialogue system testing.
Automating persona generation for user simulators can drastically reduce the labor involved in testing interview dialogue systems while enhancing the diversity of user interactions.
This paper addresses the issue of the significant labor required to test interview dialogue systems. While interview dialogue systems are expected to be useful in various scenarios, like other dialogue systems, testing them with human users requires significant effort and cost. Therefore, testing with user simulators can be beneficial. Since most conventional user simulators have been primarily designed for training task-oriented dialogue systems, little attention has been paid to the personas of the simulated users. During development, testing interview dialogue systems requires simulating a wide range of user behaviors, but manually creating a large number of personas is labor-intensive. We propose a method that automatically generates personas for user simulators using a large language model. Furthermore, by assigning personality traits related to communication styles when generating personas, we aim to increase the diversity of communication styles in the user simulator. Experimental results show that the proposed method enables the user simulator to generate utterances with greater variation.
