 |
|
Prerequisite:
Bachelor Degree (B.Sc/B.E/B.Tech) or Diploma in Computer Science, Information Technology, Linguistics, or allied streams.
Course Objective:
1. Provide an overview of the basic concepts of Artificial Intelligence with an in-depth treatment of Natural Language Processing.
2. Analyse the applicability of AI and NLP based solutions is different domains and business scenarios.
3. Illustrate different frameworks and technologies that are used to develop conversational AI applications using NLP
Course Outcome : At the end of the course students will be able to
1: Understand in-depth what is Natural Language Processing and how it works.
2: Design conversational AI solutions for various domains.
3: Build conversational AI applications for enterprises and businesses.
|
# |
Topic |
Theory |
Lab |
|
1. |
Overview of AI |
2 |
2 |
|
2. |
Approaches to AI – symbolic vs connectionist |
2 |
2 |
|
3. |
Aspects of symbolic AI |
2 |
4 |
|
4. |
Introduction to Machine Learning |
4 |
4 |
|
5. |
Common algorithms in ML |
6 |
8 |
|
6. |
Introduction to Natural Language Processing and its applications |
4 |
4 |
|
7. |
The NLP pipeline |
6 |
8 |
|
8. |
NLP Software Tools |
4 |
8 |
|
9. |
Dialogue Systems and Conversational AI |
4 |
4 |
|
10. |
Frameworks and Tools for building conversational AI systems |
4 |
4 |
|
11. |
Building a Chatbot System |
4 |
8 |
|
12. |
NLP Applications |
4 |
4 |
|
13. |
Deep Learning for NLP |
4 |
6 |
|
14. |
Industry Project |
10 |
54 |
|
TOTAL |
60 |
120 |
Course Content:
Overview of AI [2Theory, 2Laboratory]:
What is AI? – History of AI – Intelligent autonomous agents – Applications of AI – Artificial General Intelligence
Approaches to AI – symbolic vs connectionist [2Theory , 2Laboratory]:
Symbolic approach – Physical symbol system hypothesis – symbolic representation – symbolic learning – Connectionist approach – connectionist learning – connectionist representation – Integration of Symbolic and Connectionist approaches
Aspects of Symbolic AI [2Theory, 4Laboratory]:
Search – Knowledge Representation – Reasoning – Logic – Expert Systems – Planning Systems – LISP – Prolog
Introduction to Machine Learning [4Theory, 4Laboratory]:
What is machine learning? – Applications of machine learning – How does ML work? –Training data – model/algorithm – testing data – evaluation – prediction – tools required – WEKA
Common Algorithms in Machine Learning [6Theory, 8Laboratory]:
Labeled/unlabelled data – Supervised – classification – regression – Decision Tree – Conditional Random Field – Random Forest – Nearest Neighbour – Support Vector Machines – Unsupervised – clustering – association rules – Apriori algorithm – K-means
Python programming- Statements and comments – Data types and Variables – operators – Input/output – Data types – list – tuple – string – set – dictionary – arrays – matrix – File handling – file operation – directory – exceptions – Python libraries – Numpy – Pandas – Keras – scikit-learn
Introduction to NLP and its application [4Theory, 4Laboratory]:
What is NLP? – NLU and NLG – History – Computational Linguistics – Rule based approaches – Statistical approaches – Corpora – Thesauri – Gazetteers – Distributional Semantics – Word embeddings – Applications of NLP– Grammar and style checking – Document processing – Machine Translation – Dialog systems
The NLP Pipeline [6Theory, 8Laboratory]
Phonology – Morphology – Syntax – Tokenisation – Lemmatisation – Part of Speech tagging– Constituent membership – Syntactic Parsing – Named Entity Resolution– Coreference resolution – Dependency Parsing
NLP Software Tools [4Theory, 8Laboratory]:
Python tools – NLTK – spaCy – Java tools – Apache OpenNLP – Stanford NLP – Cloud based tools – IBM Watson NLU– Microsoft LUIS– AWS Comprehend
Dialogue Systems and Conversational AI [4Theory, 4Laboratory]:
History – Types of Dialogue Systems – Task oriented – Fulfillment – Chatbots – Corpus based – Retrieval based – Architecture of a Dialogue system – Dialog Manager – Intent – Entity – Slot – Turns – Initiatives – System – Mixed – Barge-in – Context – Open domain QA systems
Frameworks and Tools for building conversational AI systems [4Theory, 4Laboratory]:
IBM Watson Assistant – Google DialogFlow – RASA – LearnITyBot
Building a Chatbot System [4Theory, 8Laboratory]:
Design consideration for Chatbots – Implementing NLU – specifying intents – specifying entities – specifying slots – Selecting NLU tool – implementing the Dialog Manager – handling context – implementing the Fulfillment system
NLP Applications[4Theory, 4Laboratory]:
Sentiment analysis – Semantic Search – Document summarisation – Document auto-tagging – Latent semantic analysis (LSA)
Deep learning for NLP [4Theory, 6Laboratory]:
History – Perceptron – Activation function – Neural network – Deep neural network – Loss functions – Training – Gradient descent – DL for NLP – Distributional hypothesis – embeddings- Word2Vec – CBOW – Skip-grams
Industry Project [10Theory, 54Laboratory]:
Each student will be required to work on a Conversational AI project relevant to the industry. This will involve performing business requirements analysis, solutions design, and implementation.
Reference Books: