Natural Language Interface and Voice Flashcards
what is NLP?
Natural Language Processing
its a branch of AL concerned with the ability of a computer to be able to understand human interaction in its natural spoken or written form.
it can be defined as the ability of a machine to analyse, understand and generate human speech
what is ASR?
Automated Speech Recognition
sound input converted to text for further processing (Speech-to-Text)
what is the application natural language?
spam filters
answering questions
extracting information
summarsing information
what spam filter?
they use NLP to try and extract meaning from strings of text to help identify unwanted email and prevent it from entering clients in box in their email applications
what is the application answering questions?
search engines provide us with a wealth of information but rely on users being very specificwth the key words used to support web based searches
what is the application of extracting information?
with using financial market they use algorithmic trading as a means of managing investments
finanical investments are controlled primarily by technology which will evaluate news articles and extract relavent information to evaluate stock markets patterns before determining if clienrs should buy, sele or hold stocks
what is the application of summarising information?
information overload poses a problem for man digital users today
Facebook uses NLP to anaylse information on users colllected via social media to help determine their prefrences and to help determine which articles and adverts should be presented higher up in the news feed
why do NLP have trouble?
they try to understand the meaning of individual words, rather than the whole sentences or phrases
attempting to undertsnad whole phrases can be made more difficult due to the fact that many words have double meaning
what is morphology?
how words are fomed and their relationships with other words
what is syntax?
how words and sentence are put together
what is semantics?
the meaning of words and groups of words
what is pragmatics?
the context of spoken expressions
what is phonology?
the sound associated with spoken language
what is part of speech tagging (Pos)?
first step in NLP involves morphology
modern applications will apply a self-learning algorims which will tag words with multiple meanings
these applications will first determine the highest ocuring meanng for the tagged word and then use this to try and understand the functions
what is parse trees?
third step of NLP- considers semantics of a sentence
how do voice recognition systems work?
VR - refers to the combination of hardware and software systems which have the ability to decode a spoken command
its often used to operate devices or execute demands without the need to the use of peripherals such as keyboards or mice
what is required in the input of software of voice recognition?
sound card is required
while others such as smartphones have all the necessary hardware built into the system- sometimes specialised software
what are the approaches to speech recognition?
pattern matching- they use automated switchboards, user presented with questions Yes/No. the computer will analyse the input from the user and try match it with list of potential sound patterns
pattern and feature analysis- spoken input is recorded on microphone and digested using ADC. compared to a stored dictionary which can be used to identify what user has said
statistic analysis- more complex systems can take a more statistical approach to analysis of speech input- apply rules of grammar
artificial neural networks- scientists are looking at hoe they can be trained through the use of examples to recognise spoken input
advantages of NLP
no training required in the use of the application
increases accessibility of the application
if combined with speech recognition it can free users hands to complete other tasks
disadvantages of NLP
not all commands will be recognised
applications are complex to create and therefore expensive
background noise
advantages of voice recognition
no training is required in the use of the application
accurate spelling of words for users with literacy problems
for those with limited mobility it is more effective method of input
disadvantages of voice recognition
prone to interference
user may be required to speak more slowly and louder than normal
may only recognise limited range of voices
