cancel
Showing results for 
Show  only  | Search instead for 
Did you mean: 

Guides

Learn more about specific Flow features
Filter
Filter by Labels
Select any of the labels below to filter results.
Filters applied...
Sorted by:
It is better to have a bot with a narrow scope and a seamless handoff then a bot with a wide scope that lacks the handoff. Now go ahead, and nail that handoff. Chatbot Handoff - How to nail the bot to human handoff     You’ve designed your chatbot, tested it well and have done everything in your power to make it as good as possible. Still, your bot will not solve everything and having a good fallback option will make a difference between a successful chatbot and bot’s that will cause some anger issues. In this guide, we will guide you through the process of setting up well thought out handoff. We do that by splitting it up into 3 building blocks of a good handoff or takeover: Triggering the Handoff Procedure Setting up a Handoff Procedure Handing off from bot to human Before getting into specifics, as a general rule, it is better to have a bot with a narrow scope and a seamless handoff then a bot with a wide scope that lacks the handoff. 1. Triggering a handoff procedure As you might have experienced a chatbot is built by flows or pieces or conversation. Those flows and pieces of conversations can be reused in different parts of the conversation. The concept of reusing flows in different places is also recommended when working towards a handoff. A Handoff can be triggered by multiple triggers, think about these 3 triggers: Text trigger - An user that indicates that he would like to talk to an agent Unknown - If the user input is unknown to the chatbot, how do you   Handle the Unknown? Feedback - When the user indicates that the bot didn’t answer the question. Read more in   The Feedback Flow. All of these 3 Triggers can be followed by the same Handoff procedure before doing the actual Handoff!     2. Setting up a handoff procedure A decent handoff procedure will make the life of your agents a lot easier and will result in a better customer journey. The idea of this procedure is to enable the agent to answer questions as efficiently as possible with the highest satisfaction rate. At this stage the bot will gather more information and redirect the question to the right department or workflow. To do that, you can think about adding the following components into your chatbot. Required information - To answer a question you might need an email address, date, a specific order number or any other piece of information. If any of these pieces is still unknown, let the bot ask for it. When all the information is gathered, your agent will be able to solve the issue much faster.   Channels - What if you have multiple channels integrated? WhatsApp and Facebook Messenger allow you to send messages back to the user up to 24 hrs, allowing asynchronous messaging while live chat users expect an answer way faster. There is no one best answer for this one but in general live chat will have priority over asynchronous messaging. Business processes - How do you handle questions? If you have agents available, you can incorporate live chat. If there is a handoff during business hours a live agent can follow up directly or when it’s outside of business hours you can create a ticket in your CRM system to follow up at a later stage. To set up these rules and business processes within Flow,   conditions   are definitely worth looking into.   3. Handing over from bot to human The final step is to actually hand the conversation over to a human colleague. That is the place where an agent actually jumps in for assistance. At Flow we always support working from a single customer service tool where requests, tickets and chatbot handoffs will be picked up. To hand a conversation over to a customer service tool you can use the Handoff Reply. That will label the conversation as takeover required. When creating tickets you could also our Zapier integration or create custom tickets and integration by using APIs. Tip: Hand the conversation back and forth Don’t miss out on handing the conversation back from agent to bot. After the question is solved an agent can easily use a Start Survey event within the chatbot design and the bot picks up the survey with the user.
View full article
by Khoros Staff Khoros Staff Jun 19, 2021
Labels (1)
0
101
There are different situations that pause and resume the bot. Pause bot action You can add a   Pause bot   action within a flow to pause the bot. Whenever the action is called the bot will stop responding immediately. Using the plugin The most obvious one is using the   Flow Front plugin   and manually resuming or pausing by pressing the pause or resume button.   Trigger event A bot will automatically resume it's operation whenever an agent triggers a bot event using the plugin. Auto resume By default Flow will resume the bot after a number of minutes no more interactions has taken place. You can configure this within the Flow   Brain settings screen. Agent replies Another event that will cause the bot to pause is when a human agent sends a reply using Front. The bot will automatically resume after the auto-resume time period has passed. Resume on archive or delete Whenever a conversation is archived or deleted within Front, either manually or using a rule, the bot will resume responding to a customer. Make sure you have the webhook rule configured to watch all inboxes that are used to archive.
View full article
by Khoros Staff Khoros Staff Jun 19, 2021
Labels (1)
0
64
Instructions how to turn off notifications How to Enable and Disable Web Push Notification Web Push Notification will be visible in the top right corner of your screen when enabled. These notification are often used to pop up directly when a human takeover is required. With these notification you can respond in person within no time. On the other hand, when having a lot of takeovers you might want to disable them as they keep poppin up. Enable or Disable in Chrome Select Settings by clicking the 3 3dots on the right top Scroll down and click Advanced Select Content Settings within Privacy and Security Click on Notifications Find Flow Select Enable or Disable  
View full article
by Khoros Staff Khoros Staff Jun 19, 2021
Labels (1)
0
94
Switch from your bot to an agent. A bot is great for relative simple questions, marketing purposes or more complicated work flows but there will always be questions where the bot is better off by handing the question over to a human agent. This can be the case when is not trained for that specific question or if the bot is not trained well enough to understand the intent. This article will show you how this hand-off, or takeover works in 3 steps. 1. The Unknown Let's start with a new flow and call it "Takeover", "Unknown", or "Hand-off". Then drag & drop the unknown trigger onto your canvas. When the bot does not understand the question it will trigger this flow. 2. Manage expectations Use a text reply to manage expectations of your end user. Will the bot create a ticket or is there an agent on it's way to help? And what is corresponding time frame?   3. Follow up by an agent To follow up by an agent can be done in multiple ways. This depends on your current processes but in every process it is important to show your agent that there is a new takeover required. Let's discuss a couple of options. Email ticket Within in flow it is possible to capture the question of the user, their first name, email and other contact details. The captured data can then be   send by email   to your service center by using actions. Existing integrations Flow allows you to integrate with several customer service tools. Follow the steps within the integration menu to complete the integration. When the integration is complete you can simply use the action 'takeover' to notify the integrated platform. That platform will assign the chat to a general mailbox or agent depending on your rules. Are you using Front? Have a look at   setting up a hand-off work flow with Front. Other solutions Are you using other tooling? Feel free to contact us as your tooling might be on our backlog. Otherwise capture the required information of the user and combine it with an API call to push the data into you system and create a ticket.  
View full article
by Khoros Staff Khoros Staff Jun 19, 2021
Labels (1)
0
69
The differences between sending text and triggering events Send text vs. Trigger event Within Flow.ai you'll find rich UI elements such as Quick Replies, Buttons, Cards, Carousels and more. When using Quick Replies you will notice that the have several options, two of them being "Send text" and "Trigger event". This article describes the differences between "Send text" and "Trigger event"   Sending text When selecting "Send text" from the drop down menu you'll be asked for a label and for a value. The label of the button shows the text that is visible for your end-user within the button or quick reply, while the value represents the actual text value that will be send into the platform when your end-user uses the button. Make sure that the text that is being send into the platform should also be a text trigger for another flow, otherwise you'll have people being stuck in between flows. The option of send text is often used for: Increased customer interaction Faster communication Steering the user   Trigger event The "Trigger event" option can only link to events. You can have a flow starting with an event or you could have an event somewhere in the middle of a flow. This allows you to link to specific parts within your design. A great example is the feedback flow. You probably want your users to enter a feedback flow only after a question has been handled by the bot. In this scenario it might be wise to use the "Trigger event" with a label "Leave your feedback" and event "Feedback" The option of trigger event is often used for: Linking to specific parts within your design Looping of flows Guiding the user
View full article
by Khoros Staff Khoros Staff Jun 19, 2021
Labels (1)
0
103
When you start out small, viewing all conversations with your bot is fine. But when the number of messages grow, you want to filter and view only interactions that need the attention of a human agent. Using Front rules you can configure this by moving handoff conversations into a separate Team inbox. Move conversations to Handoff Inbox Using a team inbox, and Front rule, you can easily filter out any automated messages Go to the   Front settings   and add a new Empty Team inbox within the company Inboxes screen. Give the new Inbox a name:   Handoff Open the Company Rules screen and add a new Team Rule Give the Rule the name   Move to Handoff Select the   Inbound message   for the   When   setting Add the rules below: If   Inbox   is   Chatbot   (or   Flow by default) and   Has Tag   handoff Then   Move to Inbox   Handoff Make sure to add the new Handoff inbox to any existing rules. For example the   webhook rule   you configure during installation.       Hiding bot conversations When you have a lot of automated conversations, you only want to focus on the ones that need assistance from a human agent. Open the main Front window Click on the   Workspace options   icon just above Team Inboxes in the left sidebar Choose Inboxes and   unpin   the   Flow Inbox     Setting up notifications By default you will receive a notification whenever a new message is being received or sent by your bot. If you have a lot of conversations, you do not want to receive those. Fortunately Front provides advanced notification rules that tweak this behavior. Go to the   Front settings   and select   Notifications Within the   Basic   view turn all notifications off     Within the Advanced view, add a new rule that sends you notifications for a specific inbox     This article gives you some pointers into setting up and configuring Front for handoff. Using advanced rules Front allows tou to further automate assignment. More about: Installing Front with Flow Front help center
View full article
by Khoros Staff Khoros Staff Jun 9, 2021
Labels (1)
0
83
Operate the bot and trigger flows using the Front plugin Trigger Flows within Front From within Front an agent can manually start bot flows using the Flow Front plugin. Setting up events     Within the Flow designer you need to add different triggers. For each trigger you need to manually specify if they are allowed to be triggered by human agents. Triggering events     After selecting a conversation the plugin will show a dropdown with all events that are allowed to be triggered by agents. To trigger the event, select it and press the   Trigger event   button. After triggering the event, the bot will automatically resume operation. Troubleshooting If you are unable to trigger a specific event, please check the dashboard notifications. It might be you are triggering an event that breaks on for example a code action.
View full article
by Khoros Staff Khoros Staff Jun 19, 2021
Labels (1)
0
74
The following actions further help to extensively manage the flow designs:   Audience Projects Analytics Data Retention 
View full article
by Khoros Staff Khoros Staff Jun 19, 2021
Labels (2)
0
102
Audience provides an overview of all contacts interacting with your project. Contacts Contacts are related to channels. Once a person interacts they will show up in the contacts list. Segments Group contacts together using segments. Segments can be created by specifying conditions. Typically tags are a good way to build a segment of persons that for example opted in for a certain message; like a pre-order of a product. Import We provide a way to import contacts for channels that support phone numbers (MSISDN) using a CSV file. CSV File The following example demonstrates a CSV file that can be imported full name,phone number ,first name,last name,language, tag SUCCESSFULL_LEAD , tag OCTOBER_SALE ,metadata channel_whatsapp John Doe,+ 14155552682 ,John,Doe,en full name A required field. If the name is unknown, simply add   Anonymous phone number Required MSISDN. Provide the phone number in the   E.164 format E.164 phone number formatting entails the following: A + (plus) sign (replacing the International Call Prefix like 011) International Country Calling code Local Area code Local Phone number For example: +14155552682 first name Optional first name last name Optional last name language Optional two letter language code based on   ISO 639-1. tag Optional tags that can be attached to all imported contacts. You can specify as much as you need by putting next structure   tag {TAG_NAME}   to the headers of your csv file. Length of a single tag is limited by 1024 characters. metadata Optional metadata fields that can be attached to all imported contacts. You may want to use this information later in   cloud actions   or flow conditions. You can specify as much as you need by putting next structure   metadata {KEY_VALUE}   to the headers of your csv file. Length of a single metadata key/value record is limited by 1024 characters.
View full article
by Khoros Staff Khoros Staff Jun 19, 2021
Labels (1)
0
163
Channels, flows, brains, analytics and users are all grouped under a project. Flow also supports live collaboration within projects. Members Multiple people can work together and can be invited by the owner of a project.   Owner The owner of a project can add and remove project members. Invite members The invited members will receive an email with a request to join your project.   Export and import You can backup/export or import/restore your projects. When you export a project you will export all your training data, flows, intents, channels and cloud code.    
View full article
by Khoros Staff Khoros Staff Jun 19, 2021
Labels (1)
0
86
This is an overview of the current analytics. Please note analytics is in the process of being overhauled. Interactions The number of unique customer interaction "threads" per day. For channels like messenger these are equal to the number of unique users interacting with you app per day. When you have anonymous web chat this number can deviate. Interactions group messages together. They are channel and user agnostic. Messages The number of chat messages sent and received by the system. When someone sends 5 messages this would result in 1 interaction and 5 messages for that day. Sent messages A breakdown of the orgin of messages being sent back to users. System: Bot, webhooks, or other integrations User: Messages sent using the Chat app Moderator: Any other outside source replying to customers Sessions The number of "chats sessions" taking place. The give insight in the number of conversations taking place with one customer interacting. For example; when a user on Messenger sents 5 messages with 2 hours in between the first 2 and the last 3 messages , this would result in 1 interaction, 5 messages and 2 sessions being reported. The engament rate would be the average of messages sent (2,5). Users Shows the number of users interacting with the bot, broken down in total, new and active users. In the web widget individual users can't be tracked as it is an anonymous channel. Sentiment Automatically measured sentiment of user sent messages. Ranges from -100 (very negative) to +100 (very positive). Everything between -30 and +30 is pretty neutral.
View full article
by Khoros Staff Khoros Staff Jun 19, 2021
Labels (1)
0
122
With   data retention,   you can automatically remove conversation data that is resolved or inactive. Note: To enable this feature within your project settings you'll have to upgrade to an enterprise plan. Instead of adding a   resolve   action inside a flow, you can also resolve all conversations when they are no longer active, or remove data after the last time a resolved conversation was updated. The image below shows that inactive conversations will be resolved after 1 day. And after 1 day that they are resolved they will be permanently deleted.     You can do the same manually by clicking on the   Resolve   and   Run policy   buttons.   Resolve   will resolve all conversations that are not yet resolved and   Run policy   will remove all resolved conversations that confirm with the configured policy.   Event triggering You can also trigger an event after auto resolve or after agent’s resolve/archive. This is useful if you, for example, want to automatically send out a survey.    
View full article
by Khoros Staff Khoros Staff Jun 19, 2021
Labels (1)
0
122
A session starts when a message is received. Note: There might be any number of messages in one session. If during 15 minutes there are no messages then the session will be closed. So if a person sends a message again the new session will be started. With each new person’s message session renews. Opening event Opening event is available in   web widget. It is automatically triggered when the web widget is opened. Note: Opening event doesn't start a session. The session will be started when a person replies to a message that was sent by the opening event.
View full article
by Khoros Staff Khoros Staff Jun 9, 2021
Labels (1)
0
82
We strongly encourage you to learn some of the   basic concepts   of Flow. We also have a number of tutorials explaining different use cases and how to create them.     Hello world tutorial In this tutorial, we are going to build a simple “hello world” flow bot. We are going to train an intent to understand a greeting and reply with the text “hello world”.   We are going to design a simple “Hello World” bot. Create a flow Once you have logged into flow.ai, select the   Flows   view and press the add button to create a new   flow.   Add a flow by clicking on the plus icon. Add a trigger After creating a new flow rename it to   Hello World, drag and drop a   text trigger   from the sidebar and press   save.   Rename your flow to “Hello World”   Drag and drop the text trigger Train the intent After dragging and dropping the   text trigger, type in   hello.   Type in “Hello” Press the train button to add   training examples   to the intent that is connected to this trigger.   Click on train to train the intent. Rename the intent to   Intent Greeting   Rename your intent to “Intent Greeting” In order to give the bot the ability to respond to a variety of inputs, we need to add additional examples. The AI engine will start using NLP (Natural Language Processing) when it has at least 5 examples. Note: The more examples, the better the AI engine will work. We advice to add at least 10 examples for an intent to work well After adding the examples press save and return to the flow. Add a reply The bot is now trained to understand a greeting, let’s drag and drop a   text reply   to the flow. Type in   Hello World   as the text.   Drag and drop a text reply.   Reply with the text “Hello World”. Reply in context The bot understands a greeting and will reply with   Hello World. We can now reuse our trained intent to respond differently when a user has already greeted the bot. Select the library tab within the sidebar and drag and drop the “Intent Greeting” below the reply. And again place a* text reply *after that with a text like: “Hello Again!”   Drag and drop the “Intent Greeting from your library to the flow.   This is what your final flow should look like. Source You can   download the completed project   and   import   it within a Flow project
View full article
by Khoros Staff Khoros Staff Jun 9, 2021
Labels (1)
0
112
Flow versions is a backup system that enables you to easily manage your projects. You can set up versions for each Flow project. Versions include several new functionalities such as environments, drafts, development, production, versions. You can easily create and promote drafts, roll back versions, and switch between the environments.  Project Environments Environments are isolated projects. Each environment includes its flows, intents, entities, actions, integrations. There are three types of environments: Draft- Drafts are empty projects that can be used to update the existing project in the development environment. This is the initial step of creating or updating a project.  Development- The development environment is the workspace for testing an application before moving the project to production. You can delete the drafts and projects however, you can not delete the entire development environment.  Production - The production environment should be used for end-users as it is read-only. You can not edit or delete versions from production.  Version- A version is created each time the production environment is promoted. It can also be used as a backup of the production environment. 1. Open Flow.ai Dashboard and click Versions 2. Click ENABLE VERSIONS 3. Click + symbol 4. Select Empty draft 5. Choose New Draft template. In this example, we choose Blank. 6. Click NEXT STEP 7. Enter Bot or project name 8. Click CREATE 9. Click PROMOTE TO DEV 10. Click PROMOTE TO PROD 11. Click Deploy 12. Click Export to save a copy in your local system 13. That's it. You're done. Here's an interactive tutorial   https://www.iorad.com/player/1892066/Enable-Flow-Versions   Enable Project Versions  You can see the environments only after you enable Versions for a project.  Note: After you enable the Versions for a project you cannot disable it.  Sign in to the Khoros Flow dashboard.  Select a project. Access the Version tab from the left panel. Click ENABLE VERSIONS    Create a new draft  Follow the below steps to create a draft for the development environment. Click the + icon of the development project.  Select Empty draft to create a new draft. Or Select Import to add an existing project. Click Upload and browse the project. Click NEXT STEP.  In the Empty project window, enter the project details and click CREATE.  Promote to Development After the new draft is created and updated, you can promote the project to the development environment. This is an essential step for testing the project and ensuring that the project is functional without any errors.  Select the project draft from the available list. Click PROMOTE TO DEV . Promote to Production Promoting a draft to the production impacts Flow end-users. A draft can be promoted to a production environment only after it is promoted to the development environment.             Click PROMOTE TO PROD on the development environment.    You can also perform the following action for the versions that are promoted to production.    Export a project version: You can export the JSON file of a project and save it as a draft.   Rollback a project version : You can revert a project version after promoting it to the production environment.
View full article
by Khoros Staff Khoros Staff Dec 16, 2021
Labels (1)
0
205
The Payload list/button support for WhatsApp enhances the user experience for the end-users.  It helps to create preset options for the users who are seeking instant consultation over WhatsApp. The users can choose from the available options, this eases the process by reducing the typing effort for the end-users. This also curates the inputs that are sometimes not recognized by the bot.  Implementing the Payload support for WhatsApp also helps to retain and retrieve the customer resolution histories, if required. To implement payload list/ button support for WhatsApp, ensure that you have integrated WhatsApp.  Create Payload List Support for WhatsApp In a Flow, create a Trigger Event . In the Replies section, drag List and create List Items for the list. Update the following for List Items: Note:    You can add multiple lists and list items. To avoid errors, ensure that the text length in the field is in-inline with WhatsApp/ Facebook restrictions . You can see the error messages in the Notification section.  List Fields  Description Header Type   Select a reply format (text, image, video, or document).  This is set to None, by default.  Header Value Enter a List name that broadly defines the list. Note: You should select a header type in order to enter a value. Body Enter additional information about the list or customer concerns that must be addressed. Section Title Enter section name that broadly defines the list options. List Item Title Enter option names for the customers.  List Item Description Enter additional information to describe the option.  List Item Type Select the list type:  Send Text Value- To send a text reply.  Trigger Event- To trigger another event. Value or Event Depending on your selection in the above field: Enter the text reply for Send Text Value. Select an event for Trigger Event.  Note: The event needs to be created first.  PARAMS (Optional) Enter the Label and Value.  This information is archived in the database with the label and value for future retrieval. The default use case is using param in bot conditions. Button Text Enter final call to action for the customers, once the customer chooses from the list options. Footer (Optional) Enter additional information or announcements, if any.    Create Payload Button Support for WhatsApp In a Flow, create a Trigger Event . For the Replies section, add Buttons . Update the following for Button options: Note:    You can add multiple lists and list items. To avoid any errors, ensure that the text length in the field is in-inline with the WhatsApp/ Facebook restrictions. You can see the error messages in the Notification section.  List Fields  Description Header Type (Optional) Select a reply format (text, image, video, or document).  This is set to None, by default.   Header Value Enter a Button name that broadly defines the list. Note: You should select a header type in order to enter a value. Body Enter additional information about the list or customer concerns that must be addressed. Button Title Enter option name for the customers.  Button Type Select the button type:  Send Text Value- To send a text reply.  Trigger Event- To trigger another event.  Button Value Depending on your selection in the above field: Enter the text reply for Send Text Value. Select an event for Trigger Event.  Note: The event needs to be created first PARAMS  Enter the Label and Value.  This information is archived in the database with the label and value for future retrieval.  Footer Enter additional information or announcements, if any.   Separate list section items/ button options should have different Text Values or Trigger Events in order to provide distinct resolutions for the options selected by the end-users. 
View full article
by Khoros Staff Khoros Staff Nov 19, 2021
0
232
While the chatbots help to automate the conservational experiences by prompting instant resolutions, Monitoring bots are used to monitor the conservations. Conversations can be captured from various channels: social media messages, and community posts. The collected data from the above scenarios are then fed to NLP analytics to build a tightly bound, deep learning model to improve the overall bot performance.  You can use Monitoring bots to understand the customer sentiments from the tone of the conversation, gain insights about your product and its consumers.  You can also use Monitoring Bots to apply tags to route conversations to the right queue in Khoros Care using the Community tag, provide more context to the Agents using the Notes tag, and for analytics purposes. You can also prioritize the conversation using the Priority tag to specify the urgency of inbound requests.  Note that the Monitoring bots cannot be used to respond to the inbound inquiries, but only monitor an ongoing conversation to later present the data based on the Flow tags used by Flow Analytics.  Select Monitoring from the Bot Mode drop-down menu to switch from Live to Monitoring bot.  You can use Monitoring bots for the following:  Intent Detection: Match the inbound inquiries to Intent with the highest confidence. An Intent will always match if there is an exact match with one of the training examples. If no Intent can be matched, the bot will trigger the Unknown Intent .  Data Collection: Collect messages sent by the customer to your brand across all integrated channels and analyze the nature of the inquiry. This can help you to create validated concept flows for your chatbot. T agging Conversations: Apply tags based on the nature of the inquiry and the method to process the data. This helps to monitor the performance of each method by comparing the number of applied tags. You can pass these tags to Khoros Care to automatically route the conversations to the right queues. Prioritizing Conversations: Use the Priority Replies to set the priorities for the conversation that notifies the urgency of the message to Khoros Care agents. You can also use Note Replies to include additional information for the Agents.  All trigger actions that regular bots support should be available for Monitoring bots as well. Based on the requirement of your flow you can also add the following reply actions: Event Action Reset Tags Handover Resolve Note It is also possible to use Intents, Entities, Topic Matching, and Cloud Code to enhance your monitoring bot. Integrating the Monitoring Bot You can integrate the Monitoring bot in the same way as a regular bot, except you should select the Monitoring mode instead of the Live mode. This will make sure the bot does not respond, but only absorb without interrupting the conversation.  Setting up Monitoring Bot Flows In a Flow, create a trigger of text type.  Drag and drop ADD Tags from the replies section.  In the Tag Action section, select the applicable tags. You can select multiple tags using a comma.  Passing Tags to Khoros Care If you want to use tags to pass the conversations to Khoros Care, you should add the tags to the Khoros Care instance manually. Tags work differently in Flow and Khoros Care, Flow tags can have a label and a value. Khoros Care, however, has tag labels hence the tag values are not automatically passed from Flow to Khoros Care.  If you are working with dynamic tags in Flow with cloud code, you’ll have to manually add each tag label to Khoros Care. To make the whole system flexible, you can also create analytics dashboards with the tag information in Metabase and iFrame. For more information on Passing Tags, see Flow Customer Services: Other Khoros Products .
View full article
by Khoros Staff Khoros Staff Feb 5, 2022
Labels (1)
0
228
Intents help to classify the user utterance. Whenever a user utterance matches the confidence threshold of an existing intent, it triggers that intent. However, if the user utterance does not match with any of the existing intent with the required confidence, it triggers the Unknown . If it matched with the Unknown , it will add the user utterance to the suggested training examples for intents that are similar to it. Flow designers can add additional training examples to the intents.  You can choose to add or remove these training data to the intent. These training examples further help to improve the NLP and overall performance of the bot.    To know more about Flow Intents, see Natural Language Processing .  Before you begin, ensure that you enable the following settings for the flow:  Select the applicable flow.  Click on the settings and click on the Languages from the menu.  Check the Enable Training Assistance option.  Add unknown examples to the Unknown Create a Flow with an Unknown trigger.  Add a reply text for unknown user inputs.  Click on the Intent tab from the left-side panel.  Open the unknown System Intent .  Click Add Examples to add the unknown suggestions.  You can view the user utterance examples that matched the intents below the threshold for the last 30 days only.  In case the same example is matched with two different intents, you can select to add the training example only to one of the intent. 
View full article
by Khoros Staff Khoros Staff Feb 5, 2022
Labels (2)
0
202
Entity annotation helps to extract useful data from the user inputs. Information is captured based on specific entity types such as time, location, currencies, other user details, etc. You can include these entities in the training examples or intents. The messaging bot detects these entity types, extracts the values, and triggers the user action/ information to complete the flow. MARK ENTITIES  automatically annotates all the entities in the selected intents, instead of the user selecting individual entities for each training example. Mark entities implement the list of existing entities that are visible in the Entity Type tab.   In the training example, only the words that match with the keywords and the synonyms of the entities are annotated. This reduces the user’s time and effort to annotate several entity keywords and synonyms manually and improves intent prediction accuracy.  As you can translate the entities in other regional languages, this auto-triggers the Mark Entities, the existing entities are annotated in the selected languages. By default, this also archives all the marked entities that are translated into different languages. You use this to mark the entities for the selected languages to create a  diversified conversational experience. For more information on Entity Annotation, see Data Extraction . 
View full article
by Khoros Staff Khoros Staff Feb 5, 2022
Labels (2)
0
250
Topic Matching helps to predict and then classify untrained data for further topic classification. These classifications are based on topic labeling, sentiments analysis, profanity detection, and spam detection.  Topic classification organizes unstructured data as similar to the Intent classification and does not require training data. In Khoros Flow, topic matching is initiated when Intent Classification fails to match with the user utterance. Some of the typical use cases for topic matching are: Listening bots (Agent Assist) Conditional fallback routing to human agents Flow matching (in addition to intents) With Topic Matching you can find any topic regardless of being implemented in a flow as a trigger or condition unless the Topic matching option is disabled.  Enable Topic Matching When the Enable topic matching is disabled the classification system can simply skip topic classification.  Note: The Enable topic matching option is disabled, by default, for existing projects and should be turned on for new projects (languages).  In a Flow, click Configuration. Click Languages. Select the Enable topic matching option. You can also translate topics for the selected Language . Like with Intent, topics can be disabled individually. When a topic is disabled, its statements, subjects, etc are no longer used in classification. Disabling the topic matching will result in the following: The topic overview screen to be hidden The topic trigger will not be shown The topic match rule for conditions will not be visible Manage Topics  Like with intents, topics can be managed within a separate view named Topics. You can not remove a topic when it is included in a Flow or condition.  With this release, Topics can be grouped like flows and entities. Each topic should have a logical name and statements when newly created. The statement should have a rule that defaults to the This message is about . The Subject of the rule will have the same value as used for the name of the topic. A topic can have multiple different statements, each statement can have multiple rules.  If all rules of a statement match, the topic is matched with a higher Confidence score that defines the accuracy of the topic matched. With Natural Language Processing (NLP), the topic matching will have its confidence level slider. The default value should be 90%. The range is between 1 and 99%. Add Topics Like with intent recognition, you can match flows based on topic classification.  In a Flow, drag and drop the Topic trigger. Enter a logical topic name.  Click Topics in the left panel.  Click + to add a topic.  Select either This message is about , Sentiments or Custom from the Hypothesis drop-down list.  Enter the Subject .  Select the Confidence level for the topic.   Click Save . Add Conditions  A Topic can also be used to match conditions. A perfect example is the use of topic matching when combined with an Unknown . You can add flow level conditions to match the topics. You can choose Topic as a rule for conditions.  Add Translations  You can also translate topics for the selected language.   If Topic match is enabled for one language, it should be enabled for any other language as well. In other words, you cannot enable it for some specific languages.  In a Flow, click Configuration. Click Languages. Click + to add more languages. Click Translate to automatically translate the topics. Select the translation language and region.  Click Next.  Click Start Translation.  The topics will be displayed in a new language tab.
View full article
by Khoros Staff Khoros Staff Mar 28, 2022
Labels (2)
0
285