It is always important to evaluate the quality of your chatbots and conversational agents in order to know the its real health, accuracy and efficiency. Chatbot accuracy can only be increased by constantly evaluating and retraining it with new data that answers your customer's queries. 

Chatbots require large amounts of training data to perform correctly. If you want your chatbot to recognize a specific intent, you need to provide a large number of sentences that express that intent, usually generated by hand. This manual generation is error-prone and can cause erroneous results.

How can we solve it?

With artificially-generated data. Since Dialogflow is one of the most popular chatbot-building platforms, we chose to perform our tests using it.

In this blog we will discuss three ways of doing your chatbot evaluation by using:

1. real world evaluation data
2. synthetic data
3. "in scope" or "out of scope" queries

Chatbots can improve customer experience in contact centers by:

• Reducing customer wait time
• Achieving a higher customer satisfaction
• Cutting down contact center expenses and increasing productivity
• Getting to know your customer better
• Using human agents only when it is necessary

Most customer service and contact center executives are honing in on bots because they can handle large volumes of queries. Thus, their service center staff can focus on more complex tasks. As the technology behind bots has improved in terms of natural language processing (NLP), machine learning (ML), and intent-matching capabilities, companies are increasingly willing to trust them to handle direct customer interaction.

In order to reduce the chatbot training time, we rely on linguistics: training the chatbot with data that is tagged for linguistic phenomena.  We solve this problem by reducing the words in different user queries to their lemmatized form, so we can later train the bot with those terms linked to their respective inflected forms.

We all know that chatbot training requires a phase (o "some training time") before they can start interacting with users. However, taking a lot of time can be risky for businesses since its customers may go to other competitors who have their technology ready to satisfy his clients.

The training period of a conversational chatbot involves feeding the bot with different variations of all the possible user intents. For example, if you take the sentence "turn on the lights in your living room" it can be asked in different ways:

The following practices will help you design a great conversation between your chatbot and your client:

• Transparency
• Avoid using an excess of predetermined links and buttons
• Use an NLP middleware approach
• Include politeness and small talk
• Tolerate typos and slightly ambiguous formulations
• Always include domain specific terminology
• Define your bot's tone
• Keep the number of possibilities limited
• Let the user know when the bot is "thinking" or processing the query

Reducing complicated, confusing processes down to a natural conversation is potentially a huge business opportunity for anyone willing to jump headfirst and create a great user experience. Chatbots are only as smart as the words you feed them. If a bot is too rudimentary, people will lose trust in the company and will feel ignored and unappreciated. UX problems appear when the user deviates from the designed linear flow.

Stemming and lemmatization are methods used by search engines and chatbots to analyze the meaning behind a word. Stemming uses the stem of the word while lemmatization, the context in which the word is being used. We'll later go into more detailed explanations and examples.