Introducing "Forecast Utterance" for Conversational Data Science
This work addresses the problem of making forecasting tasks accessible to non-experts through conversational AI, representing an incremental step by applying existing NLP techniques to a new domain-specific concept.
The paper tackles the challenge of enabling an intelligent agent to understand users' prediction goals from natural language utterances in conversational data science, introducing the concept of Forecast Utterance and framing it as a slot-filling problem. It validates the approach using two zero-shot methods (Entity Extraction and Question-Answering) on three crafted datasets, showing their effectiveness in interpreting these utterances.
Envision an intelligent agent capable of assisting users in conducting forecasting tasks through intuitive, natural conversations, without requiring in-depth knowledge of the underlying machine learning (ML) processes. A significant challenge for the agent in this endeavor is to accurately comprehend the user's prediction goals and, consequently, formulate precise ML tasks. In this paper, we take a pioneering step towards this ambitious goal by introducing a new concept called Forecast Utterance and then focus on the automatic and accurate interpretation of users' prediction goals from these utterances. Specifically, we frame the task as a slot-filling problem, where each slot corresponds to a specific aspect of the goal prediction task. We then employ two zero-shot methods for solving the slot-filling task, namely: 1) Entity Extraction (EE), and 2) Question-Answering (QA) techniques. Our experiments, conducted with three meticulously crafted data sets, validate the viability of our ambitious goal and demonstrate the effectiveness of both EE and QA techniques in interpreting Forecast Utterances.