AI Journal 2023
AI Journal 2023
Journal
Artificial Intelligence
Semester VI
[Type text]
K. J. Somaiya College of Engineering, Mumbai-77
10. Virtual Lab experiment – POS tagging : Hidden Markov Model CO4
Subject Incharges-
______________________________________________________________________
Objective: To study various AI projects for their architectures. Complexity,
programming language, applications and other AI related concepts.
______________________________________________________________________
Expected Outcome of Experiment:
______________________________________________________________________
Books/ Journals/ Websites referred:
1. http://en.wikipedia.org/wiki/List_of_artificial_intelligence_projects
2. http://www.cs.cornell.edu/courses/cs478/2002sp/mllinks/
interesting_ai_demos_and_project.htm
3. http://homepages.inf.ed.ac.uk/rbf/AIMOVIES/AImovai.htm
4. “Artificial Intelligence: a Modern Approach” by Russell and Norving, Pearson
education Publications
5. “Artificial Intelligence” By Rich and knight, Tata McGraw Hill Publications
______________________________________________________________________
Pre Lab/ Prior Concepts:
History and evolution of AI, Artificial intelligence: definitions and theories.
______________________________________________________________________
Historical Profile:
AI research is highly technical and specialised and is also divided by several
multidisciplinary technical issues. So far there are many projects those have been
developed and are in progress to work on those issues. Students must learn the
applications of intelligent robots by studying various such projects to know the depth
and complexity of the course.
_____________________________________________________________________
New Concepts to be learned:
Applications of AI, Current research and future research potential in the field.
Somaiya Vidyavihar University 2023-24 Batch Page 2
K. J. Somaiya College of Engineering, Mumbai-77
_____________________________________________________________________
Project Description:
_____________________________________________________________________
Objective: Developing a basic level agent program that runs on condition-action rules
______________________________________________________________________
______________________________________________________________________
Books/ Journals/ Websites referred:
1. https://www.csupomona.edu/~jrfisher/www/prolog_tutorial/contents.html
2. http://www.csupomona.edu/~jrfisher/www/prolog_tutorial/pt_framer.html
3. http://www.doc.gold.ac.uk/~mas02gw/prolog_tutorial/prologpages/
4. “Artificial Intelligence: a Modern Approach” by Russell and Nerving,
Pearson education Publications
5. “Artificial Intelligence” By Rich and knight, Tata McGraw Hill Publications
6. “Prolog: Programming for Artificial Intelligence” by Ivan Bratko, Pearson
education Publications
______________________________________________________________________
Pre Lab/ Prior Concepts: Intelligent Agent, Agent Architectures, Rule base Vs
Knowledge Based approach
______________________________________________________________________
Historical Profile: Agent programs for simple applications need not be very
complicated. They can be based on condition-action rules and still they give better
results, though not always rational. The family tree program makes use of similar
concept.
_____________________________________________________________________
Action🡨RULE-ACTION [rule]
Returnaction
This approach follows a table for lookup of condition-action pairs defining all possible
condition-action rules necessary to interact in an environment.
Base Knowledgebase:
Rules:
______________________________________________________________________
Objective: To use the concepts of knowledge engineering to design and solve moderate
complex problem.
______________________________________________________________________
Expected Outcome of Experiment:
______________________________________________________________________
Books/ Journals/ Websites referred:
1. https://www.csupomona.edu/~jrfisher/www/prolog_tutorial/contents.html
2. http://www.csupomona.edu/~jrfisher/www/prolog_tutorial/pt_framer.html
3. http://www.doc.gold.ac.uk/~mas02gw/prolog_tutorial/prologpages/
4. http://classes.soe.ucsc.edu/cmps112/Spring03/languages/prolog/
PrologIntro.pdf
5. “Prolog: Programming for Artificial Intelligence” by Ivan Bratko, Pearson
education Publications
6. “Artificial Intelligence: a Modern Approach” by Russel and Norving, Pearson
education Publications
7. “Artificial Intelligence” By Rich and knight, Tata Mcgraw Hill Publications
______________________________________________________________________
Pre Lab/ Prior Concepts:
Agents, Agent Architecture, Programming with PROLOG
______________________________________________________________________
Historical Profile:
Knowledge is vast, uncertain and continuously changing. These properties of
knowledge make it difficult to arrive at a result. A murder mystery is a kind of situation
which depicts the uncertain nature of knowledge and also emphasizes the need of
choosing right clauses from entire knowledgebase to make a decision. He goal based
agent architecture and some knowledge engineering can help in solutioning of such
problems.
The logical agents are complex but they can reason and learn from the actions and new
precepts. They are less like acting and think like humans but more like acting and
thinking rational agents.
Knowledge and reasoning play a crucial role in dealing with partially observable
environments. A knowledge based agent can combine the general knowledge with
current percept to infer the hidden aspects of the current state prior to selecting actions.
_____________________________________________________________________
New Concepts to be learned:
Knowledge engineering, implementing complex agent architecture, uncertainty in
knowledge.
_____________________________________________________________________
The Knowledge Engineering Process
1. Identify the task
2. Assemble the relevant knowledge
3. Decide on vocabulary of predicates, functions and constants
4. Encode general knowledge about the domain
5. Encode description of specific problem instance
6. Pose queries to the inference procedure and get answers
7. Debug the knowledge base
Algorithm for KB-Agent:
Problem Statement:
Team Members:
1.
2.
2. Which were built in such a way that humans had to supply the inputs and
interpret the outputs?
a. Agents
b. Sensor
c. AI System
d. Actuators
Answer:
______________________________________________________________________
Objective: Comparison and analysis of uninformed search algorithms
Analyse and solve problems for goal based agent architecture (searching and
CO 2
planning algorithms).
______________________________________________________________________
Books/ Journals/ Websites referred:
1. “Artificial Intelligence: a Modern Approach” by Russell and Norving, Pearson
education Publications
2. “Artificial Intelligence” By Rich and knight, Tata Mcgraw Hill Publications
3. http://people.cs.pitt.edu/~milos/courses/cs2710/lectures/Class4.pdf
4. http://cs.williams.edu/~andrea/cs108/Lectures/InfSearch/infSearch.html
5. http://www.cs.mcgill.ca/~dprecup/courses/AI/Lectures/ai-lecture02.pdf
http://homepage.cs.uiowa.edu/~hzhang/c145/notes/04a-search.pdf
6. http://wiki.answers.com/Q/
Informed_search_techniques_and_uninformed_search_techniques
7. www.cs.swarthmore.edu/~eeaton/teaching/cs63/.../UninformedSearch.ppt
______________________________________________________________________
Pre Lab/ Prior Concepts: Problem solving, state-space trees, problem formulation,
goal based agent architecture
______________________________________________________________________
Historical Profile:
The AI researchers have come up many algorithms those operate on state space tree to
give the result. Goal based agent architectures solve problems through searching or
planning. Depending on availability of more information other than the problem
statement decides if the solution can be obtained with uninformed search or informed
search.
Somaiya Vidyavihar University 2023-24 Batch Page 18
K. J. Somaiya College of Engineering, Mumbai-77
Its fact that not all search algorithms end up in giving the optimal solution. So, it states
the need to have a better and methodological approach which guarantees optimal
solution.
______________________________________________________________________
New Concepts to be learned: Uninformed (blind) search, iterative deepening, greedy
best first search, A* search
______________________________________________________________________
Uninformed searching techniques:
State-space tree :
2. Optimality of BFS is
a. When all step costs are equal
b. When all step costs are unequal
c. When there is less number of nodes
d. Both a & c
Answer:
______________________________________________________________________
Objective: Comparison and analysis of informed search algorithms
Analyse and solve problems for goal based agent architecture (searching and
CO2
planning algorithms).
______________________________________________________________________
Books/ Journals/ Websites referred:
1. “Artificial Intelligence: a Modern Approach” by Russell and Norving, Pearson
education Publications
2. “Artificial Intelligence” By Rich and knight, Tata Mcgraw Hill Publications
3. http://people.cs.pitt.edu/~milos/courses/cs2710/lectures/Class4.pdf
4. http://cs.williams.edu/~andrea/cs108/Lectures/InfSearch/infSearch.html
5. http://www.cs.mcgill.ca/~dprecup/courses/AI/Lectures/ai-lecture02.pdf
http://homepage.cs.uiowa.edu/~hzhang/c145/notes/04a-search.pdf
6. http://wiki.answers.com/Q/
Informed_search_techniques_and_uninformed_search_techniques
______________________________________________________________________
Pre Lab/ Prior Concepts: Problem solving, state-space trees, problem formulation,
goal based agent architecture
______________________________________________________________________
Historical Profile:
The AI researchers have come up many algorithms those operate on state space tree to
give the result. Goal based agent architectures solve problems through searching or
planning. Depending on availability of more information other than the problem
statement decides if the solution can be obtained with uninformed search or informed
search.
Its fact that not all search algorithms end up in giving the optimal solution. So, it states
the need to have a better and methodological approach which guarantees optimal
solution.
______________________________________________________________________
New Concepts to be learned: Heuristic, Informed search, greedy best first search, A*
search
______________________________________________________________________
Informed searching techniques
● A*
State-space tree :
Properties of A* algorithm:
2. A* algorithm is based on
a. Breadth-First-Search
b. Depth-First –Search
c. Best-First-Search
d. Hill climbing.
e. Bulkworld Problem.
Answer:
______________________________________________________________________
Objective: Implementation of Alpha-Beta Pruning algorithm
Analyse and solve problems for goal based agent architecture (searching and
CO2
planning algorithms).
______________________________________________________________________
Chosen Problem:
1. Which search is equal to minmax search but eliminates the branches that can’t
influence the final decision?
a. Breadth-first search
b. Depth first search
c. Alpha-beta pruning
d. None of the above
Answer:
______________________________________________________________________
Objective: Implementation of Local search algorithm
Analyse and solve problems for goal based agent architecture (searching and
CO2
planning algorithms).
______________________________________________________________________
______________________________________________________________________
New Concepts to be learned: Constraint Satisfaction, CSP with backtracking
______________________________________________________________________
Problem chosen:
a) Forward Searching
b) Constraint Propagation
a) Search Problem
b) Backtrack Problem
c) CSP
d) Planning Problem
______________________________________________________________________
Books/ Journals/ Websites referred:
______________________________________________________________________
Pre Lab/ Prior Concepts:
A learning agent is a tool in AI that is capable of learning from its experiences. It
starts with some basic knowledge and is then able to act and adapt autonomously,
through learning, to improve its own performance.
Even Planning is an important part of Artificial Intelligence which deals with the tasks
and domains of a particular problem. Planning is considered the logical side of acting.
Everything we humans do is with a definite goal in mind, and all our actions are
oriented towards achieving our goal. Similarly, Planning is also done for Artificial
Intelligence.
______________________________________________________________________
Somaiya Vidyavihar University 2023-24 Batch Page 39
K. J. Somaiya College of Engineering, Mumbai-77
Results
Conclusion
______________________________________________________________________
.
______________________________________________________________________
Objective: Self-Learning Experiment using Virtual Lab in NLP to introduce
concepts like word analysis, word generation and building POS Tagger
______________________________________________________________________
______________________________________________________________________
Objective: Understand the Virtual Lab Experiment for tagging Parts of Speech using
Hidden Markov Model.
______________________________________________________________________
Theory: Add/Discuss the Hidden Markov Model (HMM), calculating the various
Probabilities
1
Somaiya Vidyavihar University 2023-24 Batch Page 42
K. J. Somaiya College of Engineering, Mumbai-77
Simulation Results:
Postlab Questions:
2. What is the basic design for HMM for finding out POS?