Desirable Features of a Chatbot-building Platform

Abstract: This chapter is an Introduction to the book. We analyze a lack of intelligence as a major bottleneck of current dialogue systems, briefly survey current trends, discuss how to demo a chatbot and outline the pathway towards an industrialstrength one.

Abstract: Chatbots are going to be the main tool for automated conversations with customers. Still, there is no consistent methodology for choosing a suitable chatbot platform for a particular business. This paper proposes a new method for chatbot platform evaluation. To describe the current state of chatbot platforms, two high-level approaches to chatbot platform design are discussed and compared. WYSIWYG platforms aim to simplicity but may lack some advanced features. All-purpose chatbot platforms require extensive technical skills and are more expensive but give their users more freedom in chatbot design. We provide an evaluation of six major chatbot solutions. The proposed method for the chatbot selection is demonstrated on two sample businesses - a large bank and a small taxi service.

Abstract: The temptation to be able to talk to a machine is not new. Recent advancements in the field of Natural Language Understanding has made it possible to build conversational components that can be plugged inside an application, similar to other components. These components, called chatbots, can be created from scratch or with the help of commercially available platforms. These platforms make it easier to build and deploy chatbots, often without writing a single line of code. However, similar to any other software component, chatbots also have quality concerns. Despite significant contributions in the field, an architectural perspective of building chatbots with desired quality requirements is missing in the literature. In the current work, we highlight the impact of features provided by these platforms (along with their quality) on the application design process and overall quality attributes. We propose a methodological framework to evaluate support provided by a chatbot platform towards achieving quality in the application. The framework, called Hospitality Framework, is based on software architectural body of knowledge, especially architectural tactics. The framework produces a metric, called Hospitality Index, which has utilities for making various design decisions for the overall application. We present the use of our framework on a simple use case to highlight the phases of evaluation. We showcase the process by picking three popular chatbot platforms - Watson Assistant, DialogFlow and Lex, over four quality attributes - Modifiability, Security & Privacy, Interoperability and Reliability. Our results show that different platforms provide different support for these four quality attributes.

Abstract: "Chatbot" is a colloquial term used to refer to software components that possess the ability to interact with the end-user using natural language phrases. Many commercial platforms are offering sophisticated dashboards to build these chatbots with no or minimal coding. However, the job of composing the chatbot from real-world scenarios is not a trivial activity and requires a significant understanding of the problem as well as the domain. In this work, we present the concept of Intent Sets - an Architectural choice, that impacts the overall accuracy of the chatbot. We show that the same chatbot can be built choosing one out of many possible Intent Sets. We also present our observations collected through a set of experiments while building the same chatbot over three commercial platforms - Google Dialogflow, IBM Watson Assistant and Amazon Lex.

Abstract: Building a chatbot with an iterative development process poses certain challenges for the chatbot developer. The developer is expected to produce a deployable version of the chatbot at the end of a short development cycle. Every iteration should incrementally increase the capability of the chatbot and implement a subset of overall user stories based upon a priority list, similar to any other project developed using iterative development. In this regard, commercial chatbot-building platforms offer multiple advantages to the chatbot developer, provided that the developer can map these user stories in a particular form. To do so, for every query the chatbot is expected to answer, the developer must evaluate the intention of the user. Based on the intention, the query must be processed differently, which may involve execution of some business logic. In addition, the processing of the query may require specific data items which the user must supply as part of the conversation with the chatbot. Thus, the chatbot is defined by supplying a "context" that it may encounter, and the "reaction" that must take place when the context is observed. In this work, we discuss the effects of using a platform to build a chatbot and discuss the Contextual Reactive pattern used for chatbot definition.