Work with a topic

AI-generated summary

Genesys Cloud provides comprehensive topic spotting functionality within its Conversation Intelligence Speech and Text Analytics module to analyze customer interaction transcripts and identify business-level intents such as cancellations, upsells, and escalations. Topics consist of phrases representing specific intents, and each program must have at least one associated topic defined for transcript analysis to function. The system supports two topic-spotting methods: lexical topic spotting, which identifies topics through exact or near-exact keyword matching and works best for predictable language such as compliance phrases and product names; and semantic topic spotting, which identifies topics based on meaning rather than exact wording, enabling detection of customer intent across varying word choices and phrasings. Semantic topics support up to 20 phrases per topic, allowing organizations to capture nuanced customer communications. Topic phrases impact voice transcription differently depending on language. In languages other than English and Spanish, phrases boost recognition of organization-specific terminology and adapt underlying language models. In English and Spanish, topic phrases do not influence transcription accuracy but continue to function for topic detection and analytics purposes. When detected, the system registers topics as events with specific start time, end time, name, and type attributes, displaying topic markers in interaction transcripts to highlight identified phrases. Topic configuration requires mandatory fields including a unique Topic Name, optional Description (limited to), and Tags for classification (maximum 20 tags per topic). Users can associate topics with specific programs, select from 17 supported dialects including English variants, Spanish, Dutch, French, German, Japanese, Portuguese, Hindi, Arabic, and Korean, and designate participants as External (customer), Internal (IVR, ACD, agent, conference, or voicemail), or All. Strictness settings control matching algorithm sensitivity based on phrase confidence scores (0-100), which account for word order and transcript variations. The system employs stemming functionality to recognize word variations such as different tenses or plurals as matches to root words. Mandatory phrases must be surrounded by double quotation marks during configuration, and Korean phrases require spaces between words. Phrases can be added individually or uploaded in bulk from .txt files. Genesys Cloud's AI phrase generation feature reduces manual effort by suggesting up to 10 phrases per request, limited to 100 requests daily per organization without consuming AI tokens. Generated phrases remain inactive until administrators review, approve, and publish them. Topics are accessed through Menu > Conversation Intelligence > Speech and Text Analytics > Topics, where administrators can create new topics, filter using unpublished status, tags, and dialects, and search for specific topics. Topics can be copied for reuse and permanently deleted, which automatically removes them from all linked programs. The platform implements operational limits including a standard limit of 500 topics that can be increased to 5,000 topics upon request (an irreversible change), a dialect selection limit of 15 options, and a restriction that individual topics cannot be added to more than 10 programs through the Topic Editor page, though the Program Editor page permits this functionality. Topic spotting analyzes only the first 72, of interactions. When topic counts exceed 500, the Replace and Merge process hides the topics table and deletion confirmation text, and the Filter by Tag feature becomes unavailable. Topics are published from either the Topics list (selecting up to 50 topics at once) or individual topic pages, with the system updating all associated language models and programs upon publication. Access to speech and text analytics results requires the Speech and Text Analytics > Data – View permission. During multi-agent conferences or consults, identified topics are associated with the first agent regardless of which participant spoke the phrase. Genesys Cloud offers out-of-the-box pre-configured topics as starting options before creating custom topics, and topic markers in interaction transcripts provide access to audio information for employee training, feedback, and business problem identification

PrerequisitesClick to expand

Programs, topics, and phrases requirements
Define a program for the flow, queue, or default program. If you do not define a program, topic spotting does not analyze transcripts.
Add at least one topic to a program. If you do not define one or more topics for a program, topic spotting does not analyze transcripts.
To allow agents to view speech and text analytics results, grant the Speech and Text Analytics> Data – View permission. Without this permission, agents cannot view any speech and text analytics including voice transcripts, topics, sentiment and agent empathy, categories, and acoustics data.

A topic is made up of phrases that represent a specific intent; for example, cancellation. Each program is associated with one or more topics.

Topic phrases for different languages

Depending on the language you use, topic phrases can boost voice transcription accuracy. However, in English and Spanish languages, topic phrases do not impact voice transcription.

Languages other than English and Spanish

Topic phrases help boost the recognition of specific words and phrases in voice transcription. They adapt the underlying language models to better recognize organization-specific terminology used in conversations.

English and Spanish languages

Topic phrases do not influence voice transcription accuracy. Topics continue to function for topic detection and analytics purposes, but they are not used to adapt the transcription process.

Work with topics

A topic encapsulates a business-level intent detected within interactions. Business level intents could be a call reason (for example, cancellation), an agent skill (for example, upsell attempt), or an indication of customer dissatisfaction (for example, escalation). Customers and agents have numerous ways to signal a business intent based on the words and phrases they say. As a result, a topic contains a set of phrases that indicate that business level intent. For example, if you want to identify interactions in which the customer wants to cancel an account, you could create a topic named cancellation with the following phrases:

“close out my account”
“I need to cancel”

When you include a topic in a program, the system searches for all phrases that the topic definition includes, in all the interactions that relate to the program. When one of the phrases is found, it is identified as an event and the topic is registered as found at a specific time during the interaction. Each identified event in an interaction has a start time, an end time, a name, and a type. For example if a topic called “Loan Officer” is identified by the phrase “can offer you a loan,” the start time is the second that the agent began to say “can.” The end time is when the agent finishes saying “loan.”

The more phrases that you add to a topic, the more useful the topic. For more information, see Work with a phrase.

After you create a topic, you can modify the topic based on the accuracy of the topic results.

Notes:

Define topics before you add them to a program. However, you can change their contents after you add them to a program. The topic contents that are defined when the program is applied are the contents that the system searches for in interactions associated with the program.
Surround words and phrases in double quotation marks (“) to ensure that they are mandatory during topic detection.
During a consult or conference, when multiple agents are on a call, it is not possible to distinguish what each agent said. Therefore, any topics identified during a conversation, regardless of which agent was speaking, are associated with the first agent in the conference.
When adding Korean phrases, you must add a space between words. If a phrase consists of only one word, you do not need more space before it. Adding a space between words ensures proper readability and adherence to Korean writing conventions.
Topic limits:
- Topic spotting is performed only on the first 72,000 characters of an interaction (excluding spaces). Any content beyond this limit is not analyzed.
- Standard limit: 500 topics.
- Upon request, the limit can be increased to 5000 topics.
- Once the limit is increased to 5000 topics, it is not possible to revert the topics limit to 500 topics.
- The limit for specific dialects that you can select for filtering in the Topic editor page is 15.
- In the Topic Editor page, you cannot add the same topic to more than 10 programs. You can only add a specific topic to more than 10 programs in the Program Editor page.
- When you use Replace and Merge as part of the out-of-the-box (OOTB) topics process and more than 500 topics exist, then the topics table and the text “The following topics will be deleted” does not appear.
When exceeding 500 topics, the Filter by Tag (applicable to both Topics Grid page and Program Editor page) is unavailable.
AI-generated topic phrases:
- AI generates up to 10 phrase suggestions per request.
- AI phrase generation is limited to 100 requests per day per organization.
- AI suggestions are not used for detection until you review, approve, and save them as part of the topic.

Manage topics

Use AI to generate topic phrasesClick to expand

To reduce the manual effort of creating topic phrases, administrators can use AI to generate phrase suggestions based on the topic configuration.

Note: This feature does not consume AI tokens.

When you work with a topic, AI uses the following information to suggest phrases:

Topic name
Topic description
Selected dialect
Any existing phrases that you already added to the topic
Genesys Cloud then suggests other phrases that are likely to match the same business intent.

Generate AI phrase suggestions for a topic

AI-generated phrases are suggestions only. They are not automatically published or used for detection. They become active only after you add them to the topic and save and publish your changes.

Open the topic that you want to edit in the topic editor.
Make sure that the topic has a clear name, description, and the correct dialect.
In the phrases area of the topic, use the option to generate phrase suggestions with AI.
Wait for Genesys Cloud to return a list of suggested phrases for the topic.
Review each suggested phrase:
- Accept the phrases that are relevant for the topic.
- Edit any suggestions to better match your organization’s language.
  Note: If you edit an AI-generated suggestion, the phrase is treated as a manually added phrase from that point forward.
- Discard suggestions that you do not want to use.
Save the topic.

Supported dialects for AI-generated topic phrases

AI-generated topic phrases are available for multiple dialects. AI suggestions follow the dialect that you select for the topic.

en-US (English – United States)
en-AU (English – Australia)
en-GB (English – United Kingdom)
en-IN (English – India)
en-ZA (English – South Africa)
es-ES (Spanish – Spain)
es-US (Spanish – United States)
nl-NL (Dutch – Netherlands)
fr-CA (French – Canada)
fr-FR (French – France)
de-DE (German – Germany)
de-CH (German – Switzerland)
ja-JP (Japanese – Japan)
pt-BR (Portuguese – Brazil)
hi-IN (Hindi – India)
ar-AE (Arabic – United Arab Emirates)
ko-KR (Korean – South Korea)

View and understand the topic listClick to expand

Click Admin > Quality.
Click Topics. A list of all the available topics appears.

Click Menu > Conversation Intelligence > Speech and Text Analytics > Topics. A list of available topics appears.

Click the image to enlarge.

The following information and options appear in the Topics list:

- Create Topic – To create a new topic, click the Create Topic button. For more information, see the Create a topic section in this article.
- Publish – To publish one or more topics, click the Publish button. For more information, see the Publish a topic section in this article.
- Unpublished only– Shows the number of topics that are not published. To view only a list of unpublished topics, select this check box.
- All Tags (#) – To filter the topics list to show the topics associated with the selected tags, select one or more tags.
- All Dialects (#) – To show the topics associated with the selected dialects, select one or more dialects to filter the topics list.
- Search – To find a specific topic, use the Search field.
- Name – A unique topic name.
- Description – A free text description of the topic.
- Programs – The number of programs associated with the specific topic.
- Phrases – The number of phrases associated with the specific topic.
- Dialect – Includes the dialects associated with the specific topic.
- Tags – The Tags column shows the tags associated with the specific topic. Each topic is limited to 20 tags.
- Delete Topic – Enables you to delete the specific topic. Click the x associated with the topic that you want to delete and click Yes when you are asked to confirm the topic deletion.
  Note: When you delete a topic, it is removed from all the programs it is linked to.

View a topicClick to expand

Click Admin > Quality.
Click Topics. A list of available topics appears.
Click Menu > Conversation Intelligence > Speech and Text Analytics > Topics. A list of available topics appears.
From the Name column, click the name of the topic you want to view. The selected topic details are displayed.
Review the topic details:
- Topic Name – Contains a unique name that is easy to correlate to the business issue.
- Description – Contains a free text description of the program.
To view the following information about the topic, click the Details tab. When you open a topic, this tab opens by default.
- Tags – Contains the tags associated with the specific topic. A tag enables you to classify and group programs and topics. Click the x next to the tag if you want to remove the tag from the specific topic.
- Programs – Contains the number of programs associated with the selected topic. To see the selected programs, select or clear selected programs, and search for specific programs, click the Programs field. For more information, see Work with a program.
- Strictness – Indicates how strict the matching algorithm is when matching phrases in a transcript. The matching algorithm views the phrase confidence after stemming to determine if the phrase is a match. For more information, see How do I choose appropriate strictness values when entering a topic phrase?
  - Phrase confidence – A score between 0 and 100 that speech and text analytics generates during phrase matching. Phrase confidence depends on word order and any missing or substituted words within the transcript.
  - Stemming – The process of transforming a word to the base root of that specific word. As a result, different tenses or plurals of a specific word are considered matches. For example, when the root word is “cancel,” then “canceled,” “cancels,” and “canceling” are considered a match.
- Dialect – Indicates the dialect associated with the specific topic.
- Participants – Indicates the participant associated with the specific topic.
  - External – Indicates that the specific topic is recognized only when the phrases were said by the Customer.
  - Internal – Indicates that the specific topic is recognized only when the phrases were said by IVR, ACD, agent, conference, or voicemail.
  - Both – Indicates that the specific topic is recognized only when the phrases were said by an external participant, an internal participant, or both.
To see the list of phrases associated with the selected topic, click the Phrases (#) tab. For more information, see Work with a phrase.
- Add Phrase – To open the Add Phrase pane, click this button.
  Note: To make sure that words and phrases are mandatory, add quotation marks before and after the word or phrase.
  1. To add a single phrase, click the Single tab, enter the wording for the phrase that you want to add, select the strictness, and click Add.
  2. To upload multiple phrases from a .txt file, click the Multiple tab, upload the .txt file, and click Add.
  3. Phrase – Contains the phrase associated with the topic. To open the Edit Phrase options, click the phrase.
- Strictness – Indicates how strict the matching algorithm is when matching phrases in a transcript. The matching algorithm views the phrase confidence after stemming to determine if the phrase is a match. For more information, see How do I choose appropriate strictness values when entering a topic phrase?.
  - Phrase confidence – A score between 0 and 100 that speech and text analytics generates during phrase matching. Phrase confidence depends on word order and any missing or substituted words within the transcript.
  - Stemming – The process of transforming a word to the base root of that specific word. As a result, different tenses or plurals of a specific word are considered a match. For example, when the root word is “cancel,” then “canceled,” “cancels,” and “canceling” are considered a match.
- Delete – Enables you to delete the specific phrase. Click the x associated with the phrase that you want to delete.
- Use the Search field to find a specific phrase.
Click Close.

Create a topicClick to expand

For more information about how to create a topic from the topic miner, see Create topics from a completed miner.

Notes:

Fields marked with a red asterisk (*) are mandatory. You cannot create a program if you do not fill in the fields marked with a red asterisk.
The maximum number of tags per program and per topic is 20.

Click Admin > Quality.
Click Topics. A list of available topics appears.
Click Menu > Conversation Intelligence > Speech and Text Analytics > Topics. A list of available topics appears.
Click the Create Topic button at the top of the program list.
Configure the following topic details:
- Topic Name – Enter a unique name that is easy to correlate to the business issue.
- Description – Enter a free text description of the topic. The description is limited to 1000 characters.
- Tags – A tag enables you to classify and group programs and topics. Select existing tags or create new tags. In the field provided type the tag name (100 character limit). If the tag exists, it appears in the tag list. If you are creating a tag, the word appears with (New Tag) next to it. To add the tag, press Enter.
- Programs – Contains the number of programs associated with the selected topic. To see the selected programs, to select or clear the selected programs and search for specific programs, click the Programs field. For more information, see Work with a program.
- Strictness – Indicates how strict the matching algorithm is when matching phrases in a transcript. The matching algorithm views the phrase confidence after stemming to determine if the phrase is a match. For more information, see How do I choose appropriate strictness values when entering a topic phrase?
  - Phrase confidence – A score between 0 and 100 that speech and text analytics generate during phrase matching. Phrase confidence depends on word order and any missing or substituted words within the transcript.
  - Stemming – The process of transforming a word to the base root of that specific word. As a result, different tenses or plurals of a specific word are considered a match. For example, when the root word is “cancel,” then “canceled,” “cancels,” and “canceling” are considered a match.
- Dialect – Indicates the dialect associated with the specific topic. Select a dialect from the Dialect field.
- Participants – Indicates the participant associated with the specific topic. To select one of the following participants, click within the Participants field.
  - External – Indicates that the specific topic is recognized only when the phrases were said by the Customer.
  - Internal – Indicates that the specific topic is recognized only when the phrases were said by IVR, ACD, agent, conference, or voicemail.
  - All – Indicates that the specific topic is recognized only when the phrases were said by an external participant, internal participant, or both.
To add a list of phrases, click the Phrases (#) tab. For more information, see Work with a phrase.
Click Save Draft. A Topic successfully saved as a draft message appears in the upper right corner of the screen.
Click Close.

Edit a topicClick to expand

Note: Fields marked with a red asterisk (*) are mandatory. You cannot create a program if you do not fill in the fields marked with a red asterisk.

Click Admin > Quality.
Click Topics. A list of available topics appears.
Click Menu > Conversation Intelligence > Speech and Text Analytics > Topics. A list of available topics appears.
From the Name column, click the name of the topic you want to change. The selected topic details are displayed.
Change one or more of the topic details:
- Topic Name – Change the topic name. Enter a unique name that is easy to correlate to the business issue.
- Description – Change the free text description of the topic.
- Tags – A tag enables you to classify and group programs and topics. Select one or more existing tags and create one or more new tags. In the field provided type the tag name (100 character limit). If the tag exists, it appears in the tag list. If you create a new tag, then the word appears with (New Tag) next to it. Press Enter.
- Programs – Contains the number of programs associated with the selected topic. To see the selected programs, select or clear selected programs, and search for specific programs, click the Programs field. For more information, see Work with a program.
- Strictness – Indicates how strict the matching algorithm is when matching phrases in a transcript. The matching algorithm views the phrase confidence after stemming to determine if the phrase is a match. For more information, see How do I choose appropriate strictness values when entering a topic phrase?
  - Phrase confidence – A score between 0 and 100 that speech and text analytics generates during phrase matching. Phrase confidence depends on word order and any missing or substituted words within the transcript.
  - Stemming – The process of transforming a word to the base root of that specific word. As a result, different tenses or plurals of a specific word are considered a match. For example, when the root word is “cancel,” then “canceled,” “cancels,” and “canceling” are considered a match.
- Dialect – Indicates the dialect associated with the specific topic. Select a dialect from the Dialect field.
- Participants – Indicates the participant associated with the specific topic. Select one of the following participants from the Participants field:
  - External – Indicates that the specific topic is recognized only when the phrases were said by the Customer.
  - Internal – Indicates that the specific topic is recognized only when the phrases were said by IVR, ACD, agent, conference, or voicemail.
  - All – Indicates that the specific topic is recognized only when the phrases were said by an external participant, internal participant, or both.
To add a list of phrases, click the Phrases (#) tab. For more information, see Work with a phrase.
Click Save Draft. A Topic successfully updated as draft message appears in the upper right corner of the screen.
Click Close.

Copy a topicClick to expand

Click Admin > Quality.
Click Topics. A list of available topics appears.
Click Menu > Conversation Intelligence > Speech and Text Analytics > Topics. A list of available topics appears.
From the Name column, click the name of the topic that you want to copy. The selected topic details are displayed.
Click Copy. The following message appears:
“You successfully copied the <topic name>. The copied topic will only be created after you click Save.”
Make your changes and click Save Draft. A Topic successfully saved as draft message appears.
Click Close.

Delete a topicClick to expand

Click Admin > Quality.
Click Topics. A list of available topics appears.
Click Menu > Conversation Intelligence > Speech and Text Analytics > Topics. A list of available topics appears.
To delete the topic from the topic list, next to the topic that you want to delete, click Delete .
To delete the topic from the topic’s page, perform these steps:
1. From the Name column, click the name of the topics that you want to delete. The selected topic details appear.
2. Click Delete Topic. A Topic Editor message asks you to confirm the deletion.
Click Yes. A Topic successfully deleted message appears.

Publish a topicClick to expand

To implement new topics or topic changes, you must publish the specific topic. After you publish new or updated topics, then interactions in the database process accordingly.

You can publish a topic from the Topics list and from the specific topic page.

Publish a topic from the Topics list

Click Admin > Quality.
Click Topics. A list of available topics appears.
Click Menu > Conversation Intelligence > Speech and Text Analytics > Topics. A list of available topics appears.
Click Publish.
In the Publish Topics window, select the topics that you want to publish. You can select up to 50 topics.
Click Publish Topics. The following message appears: “The publish request for {count} programs made at {timeStamp}, was completed successfully.”

Publish a topic from the topic page

Click Admin > Quality.
Click Topics. A list of available topics appears.
Click Menu > Conversation Intelligence > Speech and Text Analytics > Topics. A list of available topics appears.
From the Name column, click the name of the topic that you want to publish. The selected topic details appear.
Click Publish. The following three messages (the third message is an example) appear.
- Topic successfully updated.
- Topic successfully published.
- The publish request for {count} programs made at {timeStamp}, was completed successfully.
Click Close.

Notes:

When you publish a topic, the system updates all associated language models and programs.
If you add a topic to an unpublished program, then the system does not update the program.

View a topic marker in the interaction transcriptClick to expand

An interaction transcript provides a clear view of the conversation between external (customer) and internal (IVR, ACD, agent, conference, or voicemail) participants and highlights what the participants say. Voice transcription helps the user overcome potential difficulties such as differences in volume, accents, timing, and connection quality. It also provides access to information stored in the audio, which can be used to improve employee training and feedback and identify business problems. Topic markers appear in the transcript when a topic phrase is identified. For more information, see About voice transcription.

View a topic marker in the interaction overviewClick to expand

View out-of-the-box topicsClick to expand

Before you create the first topic, click Admin > Quality.
Click Topics. A list of available topics appears.

Before you create the first topic, click Menu > Conversation Intelligence > Speech and Text Analytics > Topics. A list of available topics appears.

Note: When you use Replace and Merge as part of the out-of-the-box (OOTB) Topics process and more than 500 topics exist, then the topics table and the text “The following topics will be deleted” do not appear.

For more information, see Out-of-the-box topics.

Difference between lexical and semantic topic spottingClick to expand

You can configure topic spotting using either lexical or semantic methods, depending on how you want to identify topics in conversation transcripts.

Lexical topic spotting

Lexical topic spotting identifies topics based on exact or near-exact words and phrases. This method matches predefined keywords in transcripts and works best when the language is predictable and consistent.

Use lexical topic spotting when you need precise detection of specific terms, such as:

Compliance phrases
Product names
Required disclosures

Semantic topic spotting

Semantic topic spotting identifies topics based on meaning rather than exact wording. This method uses natural language understanding to detect intent, even when different words or phrasing are used.

Use semantic topic spotting when conversations vary or when customers express the same idea in different ways, such as:

Identifying cancellation intent
Detecting dissatisfaction without relying on exact keywords

Note: Each semantic topic supports up to 20 phrases.

[NEXT] Was this article helpful?

Get user feedback about articles.

Loading...

Work with a topic

Topic phrases for different languages

Languages other than English and Spanish

English and Spanish languages

Work with topics

Manage topics

[NEXT] Was this article helpful?