Language detection

  • Release version: Xanadu
  • Updated August 1, 2024
  • 2 minutes to read
    • Summarize
    Summarized using AI
    This content was generated using new OpenAI-powered functionality. Results are provided on an as is basis and are not guaranteed to be accurate or complete.

    Summary of Language detection

    The language detection feature in ServiceNow uses a pre-trained model to automatically identify the language used when customers create service cases. This automation eliminates the need for agents to manually select the language for each case, streamlining case handling and improving routing accuracy. The detected language can be added to the case record and used as a skill for routing.

    Show full answer Show less

    Key Features

    • Automatic Language Identification: Detects the language from case fields such as Short Description or Description when a case is created.
    • Integration with Task Skills: Stores detected languages as skills in the Task Skills table, enabling skill-based routing.
    • Routing Support: Enables routing of cases to agents or assignment groups with matching language skills via Advanced Work Assignment (AWA) or the assignment workbench.
    • Configurable Detection Flow: The detection process is managed by the Task Intelligence Case Language Detected flow, which triggers on case creation and uses the Dynamic Translation spoke's Detect Language action.
    • Prediction Confidence and Validation: The system evaluates the confidence of language detection results and cross-checks detected languages against the Languages [syslanguage] table to ensure accuracy before applying them.
    • Support for 20 Languages: The feature supports detection for 20 different languages, enabling broad applicability.
    • Result Storage and Monitoring: Detection outcomes are stored in the Predictor Results for Task table and the Predictor Result table, which administrators with the mladmin role can review for monitoring and troubleshooting.

    Practical Use and Configuration

    ServiceNow customers can configure the language detection model to either populate the Language field on case records or add the language as a skill to facilitate automated routing. Routing based on detected language requires separate configuration. The stored language skills allow cases to be automatically assigned to agents proficient in the detected language, improving response times and case resolution quality.

    Administrators can access detailed prediction results and manage the detection model through the Task Intelligence Admin Console. The feature integrates closely with Task Intelligence for Customer Service and leverages the Dynamic Translation spoke for language evaluation.

    The language detection feature uses a pre-trained model to detect the language used to create customer service cases. You can configure the model to add the language to the Language field on the case record and/or add the language as a skill for the case.

    Case list view with short description field text in different languages and the detected language listed in the Name column.

    Agents can save time by not having to manually select the language for each case or rely on a manual process for case routing. You can use the language skill stored in the Task Skills table to route cases to the assignment groups and agents with the necessary language skills.

    Note:
    Routing based on detected language must be configured separately.

    The language detection feature supports 20 different languages. For details, see Languages supported by Task Intelligence.

    Language detection flow

    Detecting the language used to create a case is based on a flow, Task Intelligence Case Language Detected, which is available with the Task Intelligence for Customer Service application. This flow is triggered at case creation with the condition that at least one of these case fields is not empty:
    • Short description
    • Description

    When a case is created, the system calls the Detect Language action in the Dynamic Translation spoke which evaluates the text and returns the detected language.

    If the returned language confidence is above the defined confidence threshold, the system performs a check to see if the language exists in the Languages [sys_language] table. It performs this check by matching the language code to the value in the ID field.

    If the language exists in the Languages table, the system does the following, depending on the language detection model configuration:
    • Adds the language to the Language field in the Case table.
    • Stores the prediction results in the Predictor Results for Task table [ml_predictor_results_task].
    • Stores the language skill in the Task Skills table [task_m2m_skill].
    For more information, see Create a model to detect case language.

    Prediction results

    The system stores feedback on prediction results in the Predictor Result [ml_predictor_results] table. Users with the ml_admin role can access the table and view the results.

    For language detection, the Predictor Result table stores the prediction results from the Language Detection Service spoke. The Detected Language field stores the language returned by the Language Detection Services spoke.

    The Predictor Result table also stores information about skipped and failed predictions. For more information about this table, see Components installed with Task Intelligence for Customer Service.