Ekaterina Svikhnushina

Ekaterina Svikhnushina, Pearl Pu

As conversational models become increasingly available to the general public, users are engaging with this technology in social interactions. Such unprecedented interaction experiences may pose considerable social and psychological risks to the users unless the technology is properly controlled. This highlights the need for scalable and robust evaluation metrics for conversational chatbots. Existing evaluation metrics aim to automate offline user evaluation and approximate human judgment of pre-curated dialogs. However, they are limited in their ability to capture subjective perceptions of users who actually interact with the bots and might not generalize to real-world settings. To address this limitation, we propose an approach to approximate online human evaluation leveraging large language models (LLMs) from the GPT family. We introduce a new Dialog system Evaluation framework based on Prompting (DEP), which enables a fully automatic evaluation pipeline that replicates live user studies and achieves an impressive correlation with human judgment (up to Pearson r=0.95 on a system level). The DEP approach involves collecting synthetic chat logs of evaluated bots with an LLM in the other-play setting, where the LLM is carefully conditioned to follow a specific scenario. We further explore different prompting approaches to produce evaluation scores with the same LLM. The best performing prompts, which contain few-shot demonstrations and instructions, show outstanding performance on the tested dataset and demonstrate the ability to generalize to other dialog corpora.

Ekaterina Svikhnushina, Pearl Pu

This paper explores the efficacy of online versus offline evaluation methods in assessing conversational chatbots, specifically comparing first-party direct interactions with third-party observational assessments. By extending a benchmarking dataset of user dialogs with empathetic chatbots with offline third-party evaluations, we present a systematic comparison between the feedback from online interactions and the more detached offline third-party evaluations. Our results reveal that offline human evaluations fail to capture the subtleties of human-chatbot interactions as effectively as online assessments. In comparison, automated third-party evaluations using a GPT-4 model offer a better approximation of first-party human judgments given detailed instructions. This study highlights the limitations of third-party evaluations in grasping the complexities of user experiences and advocates for the integration of direct interaction feedback in conversational AI evaluation to enhance system development and user satisfaction.

Ekaterina Svikhnushina, Pearl Pu

As chatbots are becoming increasingly popular, we often wonder what users perceive as natural and socially accepted manners of interacting with them. Some researchers maintain that humans should avoid engaging in emotional conversations with chatbots, while others have started building empathetic chatting machines using the latest deep learning techniques. To understand if chatbots should comprehend and display emotions, we conducted semi-structured interviews with 18 participants. Our analysis revealed their overall enthusiasm towards emotionally aware agents. More importantly, users' intention to accept emotional chatbots seem to hinge on how these agents respond to our specific emotions, rather than just the ability to detect human emotions. Our findings also disclosed the specific application domains where emotionally intelligent technology could improve user experience. To conclude, we summarized a set of emotion interaction patterns that inspire users' intention to adopt such technology as well as guidelines useful for the development of emotionally intelligent chatbots.

Yubo Xie, Ekaterina Svikhnushina, Pearl Pu

Open-domain dialog systems (also known as chatbots) have increasingly drawn attention in natural language processing. Some of the recent work aims at incorporating affect information into sequence-to-sequence neural dialog modeling, making the response emotionally richer, while others use hand-crafted rules to determine the desired emotion response. However, they do not explicitly learn the subtle emotional interactions captured in human dialogs. In this paper, we propose a multi-turn dialog system aimed at learning and generating emotional responses that so far only humans know how to do. Compared with two baseline models, offline experiments show that our method performs the best in perplexity scores. Further human evaluations confirm that our chatbot can keep track of the conversation context and generate emotionally more appropriate responses while performing equally well on grammar.