FAQ
Conversational bots are driving the new era of user interactions. End users don’t want to feel as if they are talking to a machine. They want to enter their queries in a free format and expect a response that provides apt information or resolution. Engati with its propriety NLP (Natural language processing) engine makes it possible to train your bot to handle such queries. Engati has further made this quite simple with the construct of using FAQs to train the bot in conversational intelligence.
Frequently Asked Questions (FAQs) are standard query response combinations that help train your bot to handle customer queries. Engati allows you to add any number of variations in each FAQ set thus helping make the training as broad as possible. For Instance, if you have a bot for a school, it would need to be trained to answer questions like what is the location of the school, what are the timings, and what is the fee. These questions with all their variations and a response for each query or intent can be uploaded as FAQs. Engati takes care of the difficult task of using that to then train the bot in handling these queries in a way that a common person may ask.
Adding and uploading FAQs are available to All Plans. The ability to configure Synonyms and Stopwords is available Business Plan onwards.
To navigate to this functionality click on the Train tab in the left panel and select the FAQ option.
Terms | Definitions |
FAQ | Frequently Asked Questions (FAQs) are the standard queries relevant to a product or a service |
Intent | Intention/Purpose of the user in the conversational flow. |
Entity | It is a data point or value which you can extract from a conversation/user query. This helps you to customize what kind of information you are collecting or how you want to associate it or want to add some custom value to it. For more information on entities and intents . |
Synonyms | These are different words that mean the same. Adding synonyms can improve FAQ responses since synonyms are assigned the same value when matching an FAQ for a valid response. |
Stopwords | Stop words are the words that are excluded in NLP while matching a given query. For more information on Synonyms and Stopwords |
NLP | Natural Language Processing is the ability of a computer program to understand human language as it is spoken. NLP is a component of artificial intelligence (AI) |
Click on the Train button and navigate to the FAQ option
There are two ways to add FAQs to your bot
One FAQ at a time: This is relevant for a scenario where the number of FAQs is limited or there are individual entries to be made from time to time.
Multiple FAQs in one go: There is an upload FAQ option where a user can upload a file containing all the FAQs. This is relevant when the bot owner has a collection of FAQs available.
Click on Add FAQ button on the FAQ page.
- Category: You can create categories for your FAQs this helps you segment relevant queries together and also helps you apply an FAQ filter in a given path. A simple example can be an organisation that has multiple departments, let’s say finance, marketing, support, and operations. You can segregate all your queries into these categories. While using the support path you can limit the FAQs to the support category.
- Language: Engati provides support for multiple languages that can be added from the Languages tab under configuration. In FAQs, you can select from the list of languages activated for your bot. The default option available here is English.
- Question: This is the query that a user asks the bot. You can add multiple questions/variations for a given response. This covers the variation that may occur when different users ask the same questions.
For example: To ask how to book a ticket 2 users may use a different query.
U1: How can I book tickets on the bot?
U2: Is there a way to make reservations on the bot?
4. Entities: When a group of values leads to the same answer you can tag entities within the FAQ rather than creating separate FAQs for all the variables.
Scenario
Query: How can I enrol in the IoT Course?
Now there may be the same procedure for enrolling in a set of subjects. Let’s say
Database Management, Artificial Intelligence, Machine learning, Data Mining, and IoT. The procedure to enrol for each of them is the same. Instead of creating 5 FAQs, you can create an Entity of type Course with the name of all the courses in it.
Creating an Entity Named Course. From Train Tab navigate to Entities and click on add entity
Now in one FAQ, you can tag that entity
- Prompt Message: This is Triggered if the query is triggered without a valid entity.
- Response Type: This can be a message as shown in the above use case or you can trigger a path in this response. A message will only provide static information while a path can lead to an interactive process.
- Answer: When the response type is a message a user can add an answer for the query/question mentioned above. This will be static information and will not change in the bot flow.
If your use case has a large number of FAQs and multiple FAQ categories you can simply upload a formatted CSV file. There is a predefined format that Engati supports for uploading an FAQ file. You can Navigate from Train to FAQ, and click on the Upload FAQ button, you will get a modal where you can upload a CSV file.
You can download the sample file which details the format required.
Instructions :
Use the FAQ CSV file upload mechanism to instantly train the chatbots with multiple FAQs. This is a structured document, the detailed usage instructions for which are provided below.
- The ResponseId field is used to tag FAQ entries. This is a system-generated value which should not be altered.
- The Action field is used to indicate what operation you are performing on the FAQ set. Valid values are – INSERT, UPDATE or DELETE
- For FAQ entries where no changes are being made, the Action field can be left blank. The ResponseId should not be modified.
- For adding new rows/entries, mark the action field as INSERT for these cases. The ResponseId field would be blank here.
- Wherever any changes are being made to an existing FAQ set, mark the action field as UPDATE for these cases. The ResponseId should not be modified.
- If an existing FAQ set is to be deleted from the bot’s training, mark the action field as DELETE for these cases. The ResponseId should not be modified.
- Use the Category field to classify FAQs and manage them better.
- Use the Response field to enter the Response of the chatbot and the subsequent columns for adding the query and query variations to denote against which user queries would the bot reply with the entered response.
- For making changes to the FAQs via CSV upload, it is recommended to download the current copy of the file and update it with the changes. It is also a good practice to eliminate data sync issues, especially when the FAQs are altered from the UI as well.
You can download all the FAQs added by you on your bot in the same format as the above file.
From Train Navigate to FAQ and click on the "..." button and select Download all FAQ.
You can check from upload history the status of your upload. This will show how many FAQs are successfully uploaded on the bot. From Train, navigate to FAQ and click on the "..." button and select View Upload History.
You can maintain FAQs in different categories. This would help you in segmenting the queries and as appropriate focus on looking for responses during a particular conversation flow/path in one or more specific categories.
A simple example can be an organization that has multiple departments, let’s say finance, marketing, support, and operations. You can segregate all your queries into these categories. While using the support path you can limit the FAQs to the Support category.
