If you want to do some prereading you can follow our blog posts. Benefits of chatbots and top 20 use cases will show you have to turn this technology into meaningful solutions. Conversational Interfaces and Chatbot posts will be your guide for understanding the concepts. For a detailed list of chatbot companies please see our guide.
There are numerous API providers in the chatbot landscape, the majority of them are focusing on Natural Language Programming (NLP) and Natural Language Understanding (NLU). It is the crucial step to decide since it will be handling the most important step in a conversational interface. But let’s start with definitions;
- Natural Language Processing (NLP): In the artificial intelligence(AI) context, NLP is the overarching umbrella that encompasses several disciplines that tackle the interaction between computer systems and human natural languages. From that perspective, NLP includes several sub-disciplines such as discourse analysis, relationship extraction, natural language understanding and a few others language analysis areas.
- Natural Language Understanding (NLU): NLU is a subset of NLP that focuses on reading comprehension and semantic analysis. The combination of NLP and NLU technologies is becoming increasingly relevant on different software areas today including bot technologies. While there are many vendors and platforms focused on NLP-NLU technologies, the following technologies are becoming extremely popular within the bot developer community.
Dialog Flow (ex API.ai) have capabilities to build speech to text and text to speech, powered by machine learning. It provides built-support for currencies and date. Supports majority of the platforms like Facebook Messenger, Slack, Alexa, and Google Assistant. It supports multiple devices ranging from laptop computers to cars. Currently supports 20+ languages. It’s free for a limited number of queries.
Wit.ai is a completely free platform including the commercial use. There are no limits on request number except they ask you to notify if you are going to exceed 1 request/sec. Supports many languages. On the other side, when your app is open, your intents, entities and validated expressions will be accessible to the community but not your logs, but still, you have the ownership of the data. Used by more than 120,000 developers. Supports not only chatbots but also wearables and home devices too.
Luis is Microsoft’s platform. It stands for Language Understanding (LUIS). A machine learning-based service to build natural language into apps, bots, and IoT devices. Quickly create enterprise-ready, custom models that continuously improve. It supports many services, but they have nice features for Azure integration.
Amazon Lex is an AWS service for building conversational interfaces into applications using voice and text. With Amazon Lex, the same deep learning engine that powers Amazon Alexa is now available to any developer, enabling you to build sophisticated, natural language chatbots into your new and existing applications. Amazon Lex provides the deep functionality and flexibility of natural language understanding (NLU) and automatic speech recognition (ASR) to enable you to build highly engaging user experiences with lifelike, conversational interactions and create new categories of products.
Watson Assistant, formerly Watson Conversation, helps you build an AI assistant for a variety of channels, including mobile devices, messaging platforms, and even robots. Create an application that understands natural-language and responds to customers in human-like conversation –in multiple languages. Seamlessly connect to messaging channels, web environments and social networks to make scaling easy. Easily configure a workspace and develop your application to suit your needs.
Natural Language Platforms (NLP) Comparison
Below table is the comparison of these five NLP platforms based on their features. However, if you still feel like you need NLP platform selection guide, feel free to check our related article.
|Allow Import/Export Model||Yes||Yes||No||Yes||Yes|
|Recognize User Intent||Yes||Yes||Yes||Yes||Yes|
|Pre-built Entries||Basic parameters||More than Basic Parameters||Huge List||Basic Parameters||Basic Parameters|
|Pre-built Intents (Domain of Knowledge)||No||Around 35 Domains||No||Around 170 intents||No|
|Save Progress through Session||Yes||Yes||Yes||Yes||Yes|
|Speech Recognition||Yes||Yes, through Google Speech||Yes||Yes, through Bing Speech||Yes, through IBM Speech to Text|
|Limits for API calls||Unlimited||Unlimited||TRIAL:|
10k text queries
5k speech queries
$0.75 per 1k queries
1k API queries/month
Unlimited API queries/month
Up to 20 work spaces
Up to 2k intents
$0.004 per speech query
$0.00075 per text query
10k API queries/month
$0.75 per 1k querries
1k API queries/month
$0.0025 per API call
Available Upon Request
|Good For||Simple B2C chatbots, MVPs||Middle Level B2C chatbots, virtual assistants, MVPs||Preview Mode, too early to judge||Cortana functionality, IoT applications, Virtual Assistants, and Chatbots||Virtual Assistants and chatbots that require IBM integration|
Courtesy of (Digiteum)
If you want to understand how to measure and test your chatbot and some of the metrics see our blog. Although, with these tools it can be quite easy to build a chatbot, but it it really hard to process natural languages and public data sometimes provide suboptimal results.
We have only scratched the surface here. To see all use cases, feel free to:
And if you have a specific business challenge,
How can we do better?
Your feedback is valuable. We will do our best to improve our work based on it.