Definitions
Artificial Intelligence
Artificial intelligence is the simulation of human intelligence, particularly with regard to the processes of learning, information gathering, research, acquisition, and logical reasoning.
ASR
Automated Speech Recognition is a technology that converts spoken language into text. The process is known as speech-to-text, automatically translating spoken words into written language.
Dialog Flow
all the possible paths that a conversation between a chatbot and a customer can take, including both customer’s statements and bot’s replies.
Intent Recognition
Intent recognition is the process of deriving meaning from a given input. In other words, intent recognition takes a given data and associates it with the target class, depending on what the user wants to achieve.
Machine Learning
Machine learning is a subfield of Artificial Intelligence used as a form of data analysis to automate analytical model building. It is based on the notion that systems can learn from data, recognize patterns, and take actions with little or no involvement from a human operator.
In short, it allows machines to simulate human intellect. As a result, using systems with artificial intelligence makes it possible to perform complicated tasks in a manner that is parallel to how people solve problems.
NLP, NLU, NLG
Natural Language Processing, Natural Language Understanding, Natural Language Generation, software systems that transform free speech into structured data (and vice versa) allowing computers to identify the meaning of the statement.
Natural Language Processing
Natural Language Processing, or NLP, is the sub-branch of Artificial Intelligence that combines linguistics, computer science, and analysis, enabling machines to interpret speech and text and provide contextually appropriate replies. Its subsets include Natural Language Understanding and Natural Language Generation.
Natural Language Understanding
Natural Language Understanding (NLU) is a subtopic of NLP that focuses on extracting context and meaning from natural language inputs. Essentially, it refers to the process through which a computer comprehends the input provided by a user.
Natural Language Generation
Natural language generation (NLG) is another subset of NLP which uses Artificial Intelligence to generate large quantities of text based on a particular dataset.
General Definitions
| Name | Definition |
|---|---|
| flow | the flow, that is the progress, of a conversation |
| intent | a goal that a user has in mind when writing or speaking to the bot |
| entity | a type of information that the bot needs to detect in a user's response |
| variable | a value that the bot saves to the memory to recall it in the future |
| phrase | training phrase, a potential user's utterance with an intent, used to train the model for classification |
| context | self-contained component of a conversation that can be reused |
| block | basic building block of a flow; several block types were pre-defined to simplify the design process |
| block edition panel | panel in which you can configure interactions in a given block; opened through the settings icon or double click on the header of a block |
| chat tester | chat window available in the Flow module, allowing for testing the conversation |
| debug mode | mode allowing to view code of the conversation and detect errors; enabled by the toggle button in the top right corner of the chat window |
| reference value | working name for a subtype of a given entity, aimed at recalling it in the flow, e.g. entity: product, reference value: mobile_phone, synonyms: mobile phone, mobile, cell phone |
| test scenarios | testable conversation scenarios //inprogress |
| knowledge base item | single entry in the Knowledge Base, having its key (name) and value (content) |
| training analytics | feature letting you test training phrases in bulk using crossvalidation; both phrases belonging and not belonging to the training set can be tested this way |
| tag | |
| Intent detector/detection (intentizer in PL) | This refers to the intent detection of an NLU model, the classification of a user’s intent. SentiOne’s intent detector (intentizer) is part of the Automate platform, a proprietary NLU model. e.g. Google’s intent detector is Dialogflow |
| tokenization | This breaks up text into smaller pieces. These smaller pieces of words and sentences are called tokens. It helps the understanding of an NLU model, allowing text to be interpreted by analyzing the sequence of words. e.g. a sentence “How are you” can be tokenized as ‘how’, ‘are’, ‘you’. |
| entity extraction or named entity extraction (NER) | This is the process of obtaining and understanding relevant data from text. It identifies data such as times, names and locations. e.g. a sentence “please remind me to call John by tomorrow” contains the entities for a person name ‘John’ and a time ‘tomorrow’. |
Updated over 2 years ago