The following terms, used within the documentation, have the meanings below:
| Term | Description |
|
Agent or Conversational agent
|
Software program that automatically converses with a user, in text or speech mode, about a specific Domain in a specific language, whether or not using natural language understanding (NLU) technologies to understand the input. |
| Anomaly | Any error or defect in operation, malfunction, shutdown, interruption, suspension, slowdown of service, system and/or application, virus, bug, inefficiency, dysfunction, lack of conformity, accident, crash, scratch and/or other anomaly, of any kind, occurring in connection with the Heres Platform and/or the Services. |
| Third-Party Channel | Channel not owned by the Committment used by the Committment to communicate with Users (e.g., whatsapp, facebook messenger, telegram, smart assistants, etc.). |
| External channels | Third Party Channels and proprietary chat of the Committent. |
| Connectors | Heres proprietary software required for integration of the Chatbot with any third-party software/application. |
| Conversation Maintenance | Set of evolutionary maintenance interventions of the conversational perimeter defined in Setup aimed at: 1. optimize the performance of the Conversational Agent in terms of comprehension using supervised machine learning techniques; 2. extend and/or modify User stories to improve the quality of user interaction and expand the Conversational Agent's conversational perimeter. |
| Conversation | Set of interactions between user and bot. The conversation opens with the user's first message and closes with a closing message from the bot that is called after a predefined (customizable) time has elapsed since the user's last message. |
| Data | Interactions and the User's personal data, as defined by the GDPR, stored on the Server. |
| Domain | Scope of chatbot application (e.g., customer service) to which specific vertical features pertain. |
| Entity | Semantic container that maps a certain data type. Together with intents, they are elements that make up the training set of the NLU engine. |
| Interaction |
Exchange of individual messages between the user and the chatbot on touchpoints or through messaging services. Each interaction consists of the ordered sequence of the following messages: - message 1: User message that can be expressed in natural language or by selection of a proposed option; - Message 2: Chatbot message (consisting of one or more balloons of textual content or in rich messages format). "Message 2" is always intended as a direct response to "message 1." Any message not traceable to this sequence is not to be considered part of an interaction. |
| Intent | User's conversational intent, i.e., the unique action to be performed to which the user's request refers. |
|
NLU or Natural Language Understanding
|
Set of techniques that enables machines to understand the structure and meaning of human language. |
|
Console operator or Operator
|
User who has access to the Conversational Agent Management Console for Human Escalation, monitoring and agent management features. |
|
Panel o Management Console |
The backoffice, which is the management console of the Heres platform, available to the Client. |
| Conversational scope | Set of User Stories related to a specific knowledge base of a Conversational Agent. |
| Knowledge Base | A term that can be used in a broad sense as a synonym for "Conversational Perimeter" and which, in a specific sense, denotes the section of the Heres console on which all User Stories managed on a given conversational agent are represented and from which bot-building features can be used. |
|
Heres platform o Platform
|
The system of software and infrastructure, designed, implemented, developed and maintained by Heres, that enables the use of the Products and the provision of the Services. In particular, it includes any material (written and non-written, documentary and non-documentary, regardless of the physical medium, if any, on which it is reproduced), element, algorithm, formula, environment, client, program (application or operating), software, hardware, interface, web page, site, computer architecture and structure, method, know-how, tool, model, application code, source code, authentication code, username, password, credential, domain name, technical, operational, functional, or any other type of instruction and/or update, that is specially conceived, created, designed, implemented, developed, or otherwise made available by the Supplier to the Customer for the purpose of and as part of the provision of the Products and Services. |
| Server | All data on the Heres platform are allocated in AWS servers located in the European Union (Ireland, Germany). |
| Messaging system | Any instant messaging software that may be used by the Principal to dialogue with users on a Third Party channel or on proprietary touchpoints. |
| Solution | The whole given by the technical platform and services provided by Heres. |
| Supervised Machine Learning | Machine learning technique of providing the computer system with complete examples to use as directions to perform the required task. |
| Touchpoint | Set of Client's digital properties (e.g., site, eCommerce, App, etc.) that host the chatbot(s). |
| Training set | Example data set to train the machine learning algorithm. In the case of the NLP engine of the chatbot it consists of a set of intents and a set of entities. |
| User Says | Single sample request associated with an intent. |
| User Story | A semantic-functional unit, built to fulfill a specific task or request from a user. A User Story can be designed to function completely autonomously or in synergy with other User Stories. |
| User | Any individual person navigating the touchpoint or using a Third Party Channel. |
| Human Escalation or Human In The Loop | Human Escalation or Human In The Loop (HITL) refers to the taking over of a conversation by a human operator, who replaces the chatbot in interacting with the user to handle a specific issue. This function can be conveyed by the "Live Chat" section of the Heres console, or by the similar section of an integrated service (e.g., Zendesk Chat). |
| Pseudo-anonymization | The processing of personal data in such a way that personal data can no longer be attributed to a specific data subject without the use of additional information, provided that such additional information is stored separately and subject to technical and organizational measures designed to ensure that such personal data is not attributed to an identified or identifiable natural person, as set forth in Article 4.5 of the GDPR. |