%0 Conference Paper %1 Zhang_2020 %A Zhang, Shuo %A Balog, Krisztian %B Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery &amp$\mathsemicolon$ Data Mining %D 2020 %I ACM %K evaluation kdd2020 recommender simulation %P 1512-1520 %R 10.1145/3394486.3403202 %T Evaluating Conversational Recommender Systems via User Simulation %U https://doi.org/10.1145%2F3394486.3403202 %X Conversational information access is an emerging research area. Currently, human evaluation is used for end-to-end system evaluation, which is both very time and resource intensive at scale, and thus becomes a bottleneck of progress. As an alternative, we propose automated evaluation by means of simulating users. Our user simulator aims to generate responses that a real human would give by considering both individual preferences and the general flow of interaction with the system. We evaluate our simulation approach on an item recommendation task by comparing three existing conversational recommender systems. We show that preference modeling and task-specific interaction models both contribute to more realistic simulations, and can help achieve high correlation between automatic evaluation measures and manual human assessments.

@inproceedings{Zhang_2020, abstract = {Conversational information access is an emerging research area. Currently, human evaluation is used for end-to-end system evaluation, which is both very time and resource intensive at scale, and thus becomes a bottleneck of progress. As an alternative, we propose automated evaluation by means of simulating users. Our user simulator aims to generate responses that a real human would give by considering both individual preferences and the general flow of interaction with the system. We evaluate our simulation approach on an item recommendation task by comparing three existing conversational recommender systems. We show that preference modeling and task-specific interaction models both contribute to more realistic simulations, and can help achieve high correlation between automatic evaluation measures and manual human assessments.}, added-at = {2022-11-02T21:42:21.000+0100}, author = {Zhang, Shuo and Balog, Krisztian}, biburl = {https://www.bibsonomy.org/bibtex/2db6590a3c5ec7512fb0e3015b42c620e/brusilovsky}, booktitle = {Proceedings of the 26th {ACM} {SIGKDD} International Conference on Knowledge Discovery {\&}amp$\mathsemicolon$ Data Mining}, description = {Evaluating Conversational Recommender Systems via User Simulation | Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining}, doi = {10.1145/3394486.3403202}, interhash = {fa6f592e9251c9e62ddfedff462afbf2}, intrahash = {db6590a3c5ec7512fb0e3015b42c620e}, keywords = {evaluation kdd2020 recommender simulation}, month = aug, pages = {1512-1520}, publisher = {{ACM}}, timestamp = {2022-11-02T21:42:21.000+0100}, title = {Evaluating Conversational Recommender Systems via User Simulation}, url = {https://doi.org/10.1145%2F3394486.3403202}, year = 2020 }