anezatra/blended-skill-talk

Blended Skill Talk

Dataset Summary

This dataset contains conversations between two personas with additional context previous utterances free messages guided messages suggestions and guided chosen suggestions allowing for the creation of natural multi-modal conversations with personality empathy and knowledge

The conversations are designed to measure a full range of technical competencies such as dialogue flow management including response times topic control and coherence of conversation It also provides a basis for exploring the impact of different conversational styles on user engagement Additionally the dataset is useful for validating distributed dialogue systems across various modalities while revealing potential biases present in different contexts Finally it enables benchmarking against similar datasets toward the development of an automatic evaluation system for assessing tactical skill talk performance over time

---

Data Structure

Fields

Field	Description
personas	List of personas participating in the conversation
additional_context	Extra context or scenario description
previous_utterance	The immediately preceding dialogue turn(s)
context	General conversation context
free_messages	Free-form user or system messages
guided_messages	Messages generated via guided prompts
suggestions	Suggested responses from ConvAI2 Empathetic Dialogues and Wizard of Wikipedia
guided_chosen_suggestions	Selected suggestions actually used in the conversation
label_candidates	Optional candidate labels null in this dataset

---

Splits

Split	Examples	Size (bytes)	Description
Train	4096	9201244	Used for model training
Validation	723	1629426	Used for validation and tuning

Total dataset size: 10830670 bytes
Total number of dialogues: 4819

---

Usage Example

from datasets import load_dataset

ds = load_dataset("anezatra/blended-skill-talk", split="train")
print(ds[0])

References

Smith, E. M., Williamson, M., Shuster, K., Weston, J., & Boureau, Y. L. (2020). Can You Put it All Together: Evaluating Conversational Agents' Ability to Blend Skills. arXiv preprint arXiv:2004.08449. (https://arxiv.org/abs/2004.08449)