Home

vCons in Five Minutes

A vCon (Virtual Conversation Container) is a standard container for storing and managing conversational data (such as transcripts and multimedia files) generated in business communication, particularly in customer-facing organizations. It contains metadata, dialog, analysis and attachments portions, which include information such as participant identification, conversation contents, sentiment analysis, context documents, and integrity checking information. vCons can be used to ease service integration, help with data management and privacy, and serve as input for communication analysis and machine learning tools. They are designed to be digital assets that are versioned and signed for data integrity and provenance. The use of vCons helps lower the cost and speed deployment of analysis tools and eases customer analysis across different product lines.

The vCon JSON object is a JSON representation of a virtual conversation, used to enable the interchange of conversation data between applications and lower layers of the network stack. The vCon has three forms: unsigned, signed, and encrypted. The unsigned form of vCon has a single top-level JSON object and is necessary for cases where the vCon is partially constructed or changes while the conversation is in progress. The signed and encrypted forms of vCon include the same top-level object. The vCon JSON object includes several keys and values such as: vcon (version of the JSON format), uuid (unique identifier for the vCon), subject (topic of the conversation), and redacted (reference to a less redacted version of the vCon).

The parties portion of the vCon is a parameter that indicates the party or parties which generated the text or recording for a particular piece of dialog. The value of parties can be expressed as a single integer, an array of integers, or an array of arrays of integers.

Single channel recordings should have a parties value of the form UnsignedInt or UnsignedInt[], where the integer value or array of integer values are the indices to the Party Object(s) in the parties array that contributed to the mix for the single channel recording. The index for the Party Object should be included even if the indicated party was silent the entire piece of dialog.

Multi-channel recordings must have a parties value that is an array of the same size as the number of channels in the recording. The values in that array are either an integer or an array of integers which are the indices to the parties that contributed to the mix for the associated channel of the recording. The index for Party Objects should be included even if the party was silent the entire conversation.

The Dialog object refers to the text, audio, or video captured from the conversation. There are two defined types: "recording" and "text". The start time of the captured dialog is stored in the "start" field, and the duration is stored in the "duration" field. The parties involved in generating the dialog are stored in the "parties" field. The "mimetype" field indicates the media type of the dialog. The original filename of the dialog can be stored in the "filename" field. The dialog content is either stored in the "body" and "encoding" fields for inline dialog, or the "url", "alg", and "signature" fields for externally referenced dialog.

Analysis Object

The Analysis Object is used to store information about an analysis data related to the conversation. The analysis data can be included inline in the JSON file, or it can be an external file. Properties

• type: This parameter is used to label the semantic type of analysis data. It is a string value and its value should be one of the following: "summary", "transcript", "translation", "sentiment", "tts".
• dialog: This parameter is used to indicate which Dialog Objects the analysis was based upon. The value of the dialog parameter is the index to the dialog or array of indices to dialogs in the dialog array to which this analysis object corresponds.
• mimetype: The media type for the included or referenced analysis file is provided in this parameter.
• filename: The filename parameter can be used to preserve the name of the file which originally contained the analysis data.
• vendor: There may not be a media type defined for the file format containing the analysis data. The vendor parameter allows the product name of the software that produced the analysis to be saved.
• schema: The schema parameter allows the data format, schema, or configuration used to generate the analysis to be saved.
• body: The body parameter is used for inline included analysis. It contains the string representation of the analysis data.
• encoding: The encoding parameter is used for inline included analysis. It contains the string representation of the encoding used for the analysis data.
• url: The url parameter is used for externally referenced analysis. It contains the URL pointing to the location of the analysis data.
• alg: The alg parameter is used for externally referenced analysis. It contains the string representation of the algorithm used to generate the signature.
• signature: The signature parameter is used for externally referenced analysis. It contains the signature for the analysis data.

Attachment Object

The Attachment Object is used to include or reference ancillary documents related to the conversation. Any file type can be included or referenced here. Properties

• type or purpose: The type or purpose property can be used to specify the semantic type or purpose of the attached file.
• party: The party parameter is used to indicate which Party Object the attachment is related to. The value of the party parameter is the index to the party in the party array to which this attachment object corresponds.