Computer Science Review 36 (2020) 100239 Contents lists available at ScienceDirect Computer Science Review journal homepage: www.elsevier.com/locate/cosrev Review article Conversational agents in business: A systematic literature review and future research directions Rodrigo Bavaresco a,* , Diórgenes Silveira a , Eduardo Reis a , Jorge Barbosa a , Rodrigo Righi a , Cristiano Costa a , Rodolfo Antunes a , Marcio Gomes a , Clauter Gatti b , Mariangela Vanzin b , Saint Clair Junior b , Elton Silva b , Carlos Moreira b a Software Innovation Laboratory (SOFTWARELAB), Applied Computing Graduate Program (PPGCA), University of Vale do Rio dos Sinos (UNISINOS), Av. Unisinos 950, São Leopoldo, Rio Grande do Sul, Brazil b Dell Inc., Av. Industrial Belgraf 400, Eldorado do Sul, Rio Grande do Sul, Brazil article info Article history: Received 8 November 2019 Received in revised form 3 March 2020 Accepted 25 March 2020 Available online xxxx Keywords: Conversational agents Chatbots Machine learning Business Industry Literature review abstract The field of business shows an increasing interest in exploring conversational agents to improve service quality and market competitiveness. Furthermore, the advances in machine learning capabilities leverage the natural language processing towards natural and straightforward dialogue experiences for industries. However, in the best of our knowledge, no literature review outlines conversational agents in the business industry, primarily taking into account computational learning capabilities. This article presents a systematic literature review that encompasses these areas looking through the use of machine learning to improve the field of business. The review followed a guideline for systematic reviews to present the literature of the last decade, emphasizing business perspectives such as domains, goals, and challenges, and computational methods for self-learning, personalization, and response generation of conversational agents. As a result, the article provides the answers of three general, three focused, and two statistical questions to address the role of artificial intelligence in conversational agents applied to business domains. In this regard, the results show that no study combines self-learning, personalization, and generative-based responses for the same business solution. Additionally, the article describes the organization of the state-of-the-art, highlighting the correlation of business perspectives and machine learning methods. The contributions of this review focus on opportunities and future research directions towards human-like conversational agents for business. © 2020 Elsevier Inc. All rights reserved. Contents 1. Introduction......................................................................................................................................................................................................................... 2 2. Related work ....................................................................................................................................................................................................................... 2 3. Materials and methods ...................................................................................................................................................................................................... 3 3.1. Research questions ................................................................................................................................................................................................ 3 3.2. Data sources and search strategy ........................................................................................................................................................................ 3 3.3. Studies selection and quality assessment ........................................................................................................................................................... 3 3.4. Data extraction....................................................................................................................................................................................................... 3 3.5. Analysis and classification .................................................................................................................................................................................... 4 4. Results.................................................................................................................................................................................................................................. 4 4.1. Research questions ................................................................................................................................................................................................ 5 4.1.1. GQ1. What are the business domains that employ conversational agents?................................................................................... 5 4.1.2. GQ2. What are the primary goals of conversational agents in business domains? ...................................................................... 5 4.1.3. GQ3. What are the future challenges of conversational agents in business domains?................................................................. 5 4.1.4. FQ1. How do the studies have considered self-learning approaches in conversational agents? ................................................. 6 4.1.5. FQ2. How do the studies have focused on personalized conversational agents? .......................................................................... 7 4.1.6. FQ3. How do the studies have considered generative-based methods in the response generation? ......................................... 8 * Corresponding author. E-mail address: rsimonb@unisinos.br (R. Bavaresco). https://doi.org/10.1016/j.cosrev.2020.100239 1574-0137/© 2020 Elsevier Inc. All rights reserved.