Titre : Conversational Search for Decision Making

Sujet proposé dans : M2 MOSIG, Projet --- M2 MSIAM, Projet

Responsable(s) :

Mots-clés : Information Retrieval
Durée du projet : 5 months
Nombre maximal d'étudiants : 1
Places disponibles : 1
Interrogation effectuée le : 02 décembre 2022, à 16 heures 12


Description

Conversational search is a method for searching for information in an interactive way [1]. The user in the conversational search engages in a natural language dialogue with the system. The system then asks the user questions to be able to specify the user’s information needs. Conversational search thus combines information retrieval with a dialogue system. 

Though conversational search can be applied on open-domain, the domain of the conversational search system is most usually limited. The goal of the conversational system created in this work is to assist the user with a decision making process [2]. In our given scenario, the user would like to make a decision whether to buy or sell the stocks. The proposed system is supposed to ask the user questions which would lead to the decision which minimizes the risk needed to be taken by the user and maximize the possible profit.

The goals of this work is to implement a conversational search system are following:

  1. Implement the conversational search system

  2. Propose the dialogue states for decision making process

  3. Propose the questions which would lead to risk minimization and maximization of the possible profit

  4. Achieve preliminary tests of the system on real users

[1] Hamed Zamani, Johanne R. Trippas, Jeff Dalton, and Filip Radlinski: Conversational Information Seeking, An Introduction to Conversational Search, Recommendation, and Question Answering, arXiv:2201.08808v1, 2022

[2] Annie Duke: Thinking in Bets: Making Smarter Decisions When You Don't Have All the Facts, Portfolio Penguin, 2018.