Thursday, 22 September 2016

AMCAT Test (Aspiration Minds Computer Adaptive Test)

What is AMCAT Test ? All the details and information about AMCAT Test.

AMCAT is the test conducted by aspring minds for testing the logical and technical abilities of a B.TECH, MBA, POST GRADUATE students. The test is all about aptitude, logical, verbal and technical (depends on their stream). This is the best test for placing in the companies, more than 3000+ companies are giving jobs based on AMCAT test.
The main problem of the companies are how to hire best employees for their dynamic company. The solution is AMCAT, it will test the skills of the student and made the company task very easy. Some companies uses their own test but many companies in india prefer AMCAT test score card.
There four steps for AMCAT test :-

  1. Register for AMCAT test on (https://www.myamcat.com/).
  2. Take the test anywhere in india.
  3. Get to know how employable are you.
  4. Get calls form companies.
The students who are going to apply AMCAT test they should know the syllabus and the pattern of the test. It onsists of four modules.

  • Quantitative Test
  • Logical Ability
  • Verbal (English Skills)
  • Optional Module (This module consists of techinical type of questions its depends on the stream of the student applying to that test).
The total time given for AMCAT test 3-hours. In this 3-hours you have to complete these four modules and the total marks for the paper is 3600 marks. Each module is for 900 (The cutoff mark is 550 +). All the 3 modules (Quant, Verbal and Logical) comes under compulsary module and optional module varies depends up on the stream of student. But u have to attempt both modules in the test. After completion of the test we have to update our profile in official website (which i had mentioned above). If you achieved above 550 you will get calls from companies and also depends on the company.
I will mention the average scores in each module :-

  • Verbal ---> 455 - 520 (average score) and above is excellent.
  • Quantitative ---> 445 - 520 (average score) and above is excellent.
  • Logical ---> 455 - 520 (average score) and above is excellent.
  • Optional module ---> 350+ (average score) and 400 above is excellent.  


Wednesday, 21 September 2016

POS Tagging Ambiguity in Natural Language Processing

Ambiguity in POS (Parts of speech) tagging :-

1st Ambiguity : -The first ambiguity is that when a word appears in the sentence the meaning of that word will differ at that case.
Example :- 
  • Mujhe Khaanaa Khaanaa hai | -------Hindi (Here we can see that khaanaa has appeared two times in the sentence but the POS tagging of the first khaanaa is NOUN and second is VERB).
  • I bank on the bank on the river bank.------English (Here we can see that bank has appeard three times in the sentence but the POS tagging of the first bank is VERB, second is NOUN, third is NOUN).
2nd Ambiguity : - For finding ADVERBS and ADJECTIVES we know the rules. We can say a word as ADVERB when that follows verb and for ADJECTIVES that should follow noun.
But,if there exists a word in between the pair of (ADVERB , VERB), (ADJECTIVE and NOUN) the machine can't give tagging to those words.
To avoid that problem of 2nd ambiguity  there are two process :-
  • First neglate the word which occurs between those above pairs.
  • Second take the list of imcomming words between them..
Process of giving POS tagging :-
  • List all possible tags in the sentence.
  • Choose the best sutiable tag for the sentence (sequence).
Example :-  People jump high.
              People--> NOUN/VERB.
              Jump---> NOUN/VERB.
              High---> NOUN/VERB/ADJECTIVE/ADVERB.

POS (Parts of Speech) Tagging in Natural Language Processing

POS is the process that attach each words of the sentence with the sutiable tags from a given set of  tags. This set of tags is called tag-set. One of the standard tag-set and commonly used tag-set is Penn-Tree tag-set. It is also a part of lexical analyzer.

  1. An important component. 
  2. Well Recognised.
  3. Phase dection, sentence formation comes after this phase of NLP.
  4. Syntatic analyser parts comes in picture after POS tagging.
For Example, RAM - It have many tags.
                NN, Noun, NUN, NM, PN, PNN, PNO.
There are many labels used for a single word. But if is having only standard label that is called Penn-Tree tag-set.
One word can be used in different sutiation with different parts of speech.
JUMP can reassembles as NOUN/VERB.
I JUMP high-----Here jump reassembles as VERB.
The last JUMP was good------Here jump reassembles as NOUN.

Penn-Tree and Tag-Set Examples :-
  • CC Coordinating Conjuation ; jack and - CC jill.
  • CD Cordinary Number ; Four - CD boys.
  • DT Deterniner ; The - DT school.
  • EX Exitential there ; There - EX was a king
  • FW Foreign word ; शादी - FW means word.
  • In proposition a subordinating conjuction ; play with - In ball.
  • JJ Adjective ; Fast_JJ car.
  • JJR Adjective ; Fastest_JJR car.
  • JJS Adjective ; superlative faster_JJS car.
  • LS list item ; 1-LS bread 2-LS butter.
  • MD model ; you may-MD go there.
  • NN Noun, singular ; water-NN is hot.
  • NNS noun, plural ; boys_NNS playing in garden.
  •  NNP proper noun, singular ; Ram_NNP went to school.
The process of Tagging : -
  1. The parts of speech tagger placed the right label with each word.
  2. The part of speech tagging is done with the help of corpus.
Corous : - It is the set of meanings like lexicon but it contain only few words which is related to the sutiation, but lexicon (dictionary) contains total words present in the language. 

Monday, 19 September 2016

Knowledge Representation in Artificial Intelligence

Knowledge Representation ;- In order to solve the complex problem encounters in artificial intelligence, one needs both a large amount of knowledge and machenism for manipulating that knowledge to create solution to new problems. It is the representation of information around which is used to solve the complex tasks. It is were important in artificial intelligence to solve complex problem. The knowledge stored in the knowledge base is represented. Artificial intelligence cannot solve the problem without storing information about world in knowledge base.

Mapping between Facts and Representation :-

  • Facts
  • Internal Representation
  • Language representation
Approaches to Knowledge Representation :-

  1. Representational Adequary
  2. Inferential Adequary
  3. Inferential Efficiency
  4. Acquisitional Efficiency
Example :- Given,

  • Akhil was a man
  • Akhil was a technocrat
  • Akhil was born in 1995
  • All men are moral
Representation of above Example :-

  • Man (Akhil)
  • Technocrat (Akhil)
  • Born (1995)
  • for all x : man (x) ---------->  moral (x)






Saturday, 17 September 2016

Expert System in Artificial Intelligence

Expert System in Artificial Intelligence :- Expert systems are the computer applications developed to solve problems in a particular domain. The system features high performance reliablity occurance and high responsive.
Capablities of Expert Systems :- It is capable of advising, instructing and assisting for demonstrative for deriving the solution for explaining interpting input for prediction results, for conclusions, for suggesting alternative options for feed back. An expert system is incapable of for substituting human decisions, processing human capablities, providing accurate results a output for in adquate information, refining of its own knowledge.
Components of an Expert System :-

  • Knowledge Base
  • Interface Engine
  • User Interface
Knowledge Base :- It contains domain specifications and high quality knowledge, it is required to exibit intelligence, the success of the component and presis knowledge .

Knowledge  :-  The data is  collection of  facts and figures, the information in the organisation data and combining data, information and experiance gives. Knowledge, a particular term is used for specific task.

How can we replace Knowledge?
An artificial intelligence the method used for organizing the knowledge in the knowledge phase is in the form of if,then,else rules.

Knowledge Acqurition :- For success of any expert system various factors ate there named as quality, completeness and accuracy of information. The knowledge phase formed by various experts, scholars and engineers vary from learning skills to easy projects. If, then, else rules are used by interferance machine which is monitor by knowledge engineers.

Interface Engine :- In this block of expert system use of efficient procedures and rules are done by the interface, it is correct in every manner if accuaries and manipulate the knowledge from knowledge phase to provide a particular solution.
In it we have strategies to work :-

  • Forward Chaining
  • Backward Chaining
User Interface :- It provides interaction between the user and the expert system, it is generally natural language processing to be used by the user in varioud tasks, in this user interface the epert system requires some particular tasks.

  • Natural language to be displayed on the screen.
  • Verbal narration in natural language.
  • Listing of rules, numbers and others necessary features on the screen.
           

A* Algorithm in Artificial Intelligence

A* Algorithm : - In best first search we found a heuristic value called evaluation function value. It is the value that estimate how far a particular node is from the goal node. A part from the evaluation function values one can also bring in cost function. Cost function indicates how much resources like time, energy, money ect... have been spend in reaching a particular node from the start node.
While evaluation function value deal with future, cost function value deal with past.
If if is possible for one obtain the evaluation function and cost function value, then A* algorithm can be used.
The sum of evaluation function value and cost function value is known as fitness.
The fitness number is the total evaluation value and the cost function value. While best first search uses the evaluation function values only for expanding the best node. Where as A* uses fitness number for its computation.

Algorithm :-

  • Step 1 :-  Put the initial node on the list BEGIN.
  • Step 2 :-  If (BEGIN = EMPTY) or (BEGIN = GOAL) terminate search.
  • Step 3 :- Remove the first node from BEGIN call this node as A
  • Step 4 :- If (A = GOAL) terminate the search process.
  • Step 5 :- Else if node has a successor than generate all of them estimate the fitness number of successor by totally the evaluation function value and cost function value sort the list by fitness number.
  • Step 6 :- Name the list as BEGIN 1
  • Step 7 :- Replace BEGIN with BEGIN 1
  • Step 8 :- GOTO Step 2.


Tuesday, 13 September 2016

Best First Search in Artificial Intelligence

Best First Search :- Best First Search was described by Judea Pearl. It is the part of  heuristic search.

Algorithm :-

  • Step 1 :- Put the initial node on the list BEGIN.
  • Step 2 :- If (BEGIN is empty) or (BEGIN is GOAL) terminate search.
  • Step 3 :- Remove the first node from BEGIN call this node as A.
  • Step 4 :- If (A is GOAL) terminate search.
  • Step 5 :- Else if node A has successor generate all of them. Find out how far they are from the goal node. Sort all the children generated so, far by the remaining distance from the goal.
  • Step 6 :- Name the list as BEGIN 1
  • Step 7 :- Replace BEGIN with BEGIN 1
  • Step 8:- GOTO Step 2 

Monday, 12 September 2016

Hill - Climbing Algorithm in Artificial Intelligence

Hill - Climbing Algorithm :- It is also known as discrete optimization algorithm.It is a iterative algorithm and part of heuristic search category.

Algorithm :- 
  • Step 1 :- Put the initial node on the list BEGIN
  • Step 2 :- If  (BEGIN is EMPTY) or (BEGIN is GOAL) terminate search
  • Step 3 :- Remove the first node from BEGIN call this node as A
  • Step 4 :-If (A = GOAL) terminate the search process
  • Step 5 :-Else if node has a successor, generate all of them.Find how far they are from the goal node sort them by the remaining distance from the goal node and add them to the beginning of BEGIN
  • Step 6 :-GOTO  Step 2
Hill climbing technique is used in some activities in day to day life.

Examples :- 
  • While listening, playing flute on transistor so that the volume controls are adjusted in a way make the music melodious. 
  • While tuning carborator of scooter the accelerator will be raised to its maximum and carborator is used so that the engine keeps on running for a long period of time.
  •  An electronic expert while making the transistor for the first time tune the radio set and mid afternoon when signal is weak.
Problems Faced(Caused) in Hill Climbing Search :-
  1. Local Maximum
  2. Platern
  3. Ridge
  • Local Maximum :- A state thet is better than all it neighbours but not so bond compare with to states that far away.
  • Platera :- It is a large area of search state in which neighbours have the same values.
  • Ridge :- It is described as a long and narrow streach of evaluated ground or a narrow elivation aa based running along or a closed surface. According to oxford english dictionary ridge means - It is a area in the path which must be travels very carefully because movement in any direction might maintain one at the same level or resultant fast descant.
How to avoid these problems in Hill Climbing search :- 
  • For Local Maximum :- Backtracking helps in undoing what has been done so far and permits to try a totally different path to attain the global peak.
  • For platern :- A big jump is the solution to the escape from the platern. It is recommended because all neighbouring points have the same values.
  • For Ridge :- To overcome the problem of ridge we are trying different paths at the same time. 

  

Heuristic Search in Artificial Intelligence

Heuristic Search :- In the previous algorithm which are totally rude in nature because they do not have any domain specific knowledge.
Brute force search is commonly used to search procedure which explore all the alternatives during search process. They do not have any domain specific knowledge, All they need are the initial, final states and a set of legal operators. Heuristic search approximation used to minimize the searching process. Generally two categories of problems use heuristic.

  1. Problem for which no exact algorithm are known and one needs to find an approximation and satisfaction solution.                                                                                                                       Example :- speech recognation, pattern recognation, natural language undestanding, computer vision.
  2. Problem for which exact solution are known but computational inflexible                                    Example :- chess, rubix cube

Search Process in Artificial Intelligence

Search process in artificial intelligence is very important concept in concept :-

  1. Brute Force Search
  2. Heuristics Search
Brute Force Search -

  • DFS - Depth First Search
  • BFS - Breadth First Search
Heuristics Search : -

  • Hill Climbing
  • Best First Search
  • Constraint Satisfaction
  • Mean - End Analysis
  • A* Algorithm 
  • AO* Algorithm
Brute Force Search  :- Brute force search is un-unformed or blind search. It is used for searching small instances. It is direct approach to solve the problem accoding to given problem statement and defination of concept involved. It is very simple and easy process to solve the small size problem. It requires n-1 multiplications.

DFS (Depth First Search) :- DFS is the search used for searching graphs and trees in data structure. The search starts with the root node and search continous to the ending of the graph or tree according to our requirement.It also explores the branches and their nodes as well. Here the data structure to store the nodes is STACK

Algorithm: -

  • GOTO the un-visited adjecent node and mark it as visited and push into stack
  • If the visited node having no child node the pop it out of the stack
  • If the visited node having child nodes the GOTO the child nodes.
  • Repeat this step untill the graph ot tree become empty
  • Remember the top of the stack will changing while we are traversaling the tree or graph
BFS (Breadth first search) :- BFS is the search used for searching graphs and trees in data structure. The only difference between DFS and BFS is that in DFS we will visit all child nodes (left and right) then we will move to the next, but in BFS we will goto the depth of the visited node (i.e if it visit the left node it will complete traversal left side of the tree and the it comes to the right node, vice versa)
Here the data structure to store the nodes is QUEUE.

Algorithm :-

  • Here there is the concept of CURRENTLY WORKING NODE
  • GOTO the root node and mark it as the currently working node 
  • Then visit the child nodes and queue it into queue data structure one by one
  • Then update the currently working node as any one of the child and de-queue it from queue data structure one by one
  • Then traversal to he childs of the currently working node and repeat the same process until the tree or graph gets finish.  

Sunday, 11 September 2016

Turing Test in Artificial Intelligence

Can machine thinks ?
In 1950  Alan Turing published an article in the mind magazine which triggered a controvercy topic can machine game which was later modified to turing test.

  • Room A (Human or Machine)
  • Room B (Human or Machine)
  • Room C (Interrogator)
In the imitation games players are 3-humans -male,a female and interrogator. The interrogator who is shielded from other too, ask questions to both of them and based on their type written answer, who is male and who is female. The aim of the male is to imitate the female and recieve the interrogator and the role of female is to provide the information that she is female.

The turing test is the slight variation of the invitation games, Where the human interrogrator now communicate with the male via a teletype- 1 human and the other a machine.

Turing propose that if laymen interrogator in Room C is not able to identify, who is in Room A or in Room B, then the machine parser intelligence. However the test is not easy as proposed, Human as far more superior, common-sence and reasoning.
If any questons put on these topics are excel. On the numerical computation part machine are accurate and faster.
  

Problem Space in Artificial Intelligence

A problem in space in that can be generated by the application on any combination of objects the problem in space contains one or more

  • State Space :- A state space is the set of all states reachable from initial state.
Example :- Water-Jug Problem
There are two jugs one is 4 gallon and another is 3 gallon jugs one neither have any meaning maker on it there there is pump that can be user to fill the jugs with water how can you get exactly two gallons of water into the four gallon jug. 
                                      State Space
                                       (x.y)        x---->4 gallon jug
                                                       y----->3 gallon jug
x=0,1,2,3,4
y=0,1,2,3

initial state (0,0)
goal state (2,n)

Production Rules for Water-Jug Problem :-
  • (x,y) ---> (4,y)                                                  Fill 4 gallon jug
  • (x,y) ---> (x,3)                                                   Fill 3 gallon jug
  • if x>0 ,(x,y) ---> (x-d,y)                                    Pour some water out of 4 gallon jug on the ground
  • (x,y) ---> (x,y-d)                                                Pour some water out of 3 gallon jug on the ground
  • if x>0, (x,y) ---> (0,y)                                        Empty the 4 gallon jug on the ground
  • if y>0, (x,y) ---> (x,0)                                        Empty the 3 gallon jug on the ground
  • if (x+y)>=4 and y>0, (x,y) ---> (4,y-(4-x))        Pour water from 3 gallon jug into 4 gallon jug until it fulls
  • if (x+y)>=3 and x>0, (x,y) ---> (x-(3-y),3)       Pour water from 4 gallon jug to 3 gallon jug
  • if (x+y)<=4 and y>0, (x,y) ---> (x+y,0)            Pour all the water from 3 gallon jug to 4 hallon jug
  • if (x+y)<=4 and x>0, (x,y) ---> (0,x+y)            Pour all the water from 4 gallon jug to 3 gallon jug
  • (0,2) ---> (2,0)                                                   Pour the 2 gallon water from 3 gallon jug to 4 gallon jug
  • (2,y) ----> (0,y)                                                  Empty the 2 gallon jug in the 4 gallon jug
Solution Set :-

4 Gallon jug                       3 Gallon jug                        Rules Applied
        0                                        0                                         Rule 2
        0                                        3                                         Rule 9
        3                                        0                                         Rule 2
        3                                        3                                         Rule 7
        4                                        2                                      Rule 5 or Rule 11
        0                                        2                                       Rule 9 or Rule 11
        2                                        0                                      Rule 9  or Rule 11     
                         




Artificial Intelligence

It is the study of how to make computer to think which at the moment people do better. The training of the machine (Computer) is also known as artificial intelligence. Now-a-days artificial intelligence becoming popular in the world, due to artificial intelligence many things in the world becoming easy and convinent.
It is divided into three tasks:-
  1. Mundane Task
  2. Formal Task
  3. Expert Task
Mundane Task :-
  • Perception
  • Natural Language 
  • Common Sence Reasoning
  • Robot Control
Formal Task :-
  • Games
  • Mathematics
Expert Task :-
  • Engineering 
  • Scientific Analysis
  • Medical Analysis
  • Financtial Analysis
Problem, Problem Space and Search
To build a system to solve a particular problem we need tools forthing
  1. Define the problem precisely
  2. Analyse the Problem
  3. Isolate and represent the task knowledge
  4. Choose the best problem solving techniques and apply it to the particular problem.

Tuesday, 6 September 2016

Top Down Parsing and Bottom Up Parsing in NLP (Natural Language Processing)

Top Down Parsing:- This parsing begins with the starting symbol and proceeds by breaking the sentence into its constituents words. It starts with the symbol 'S' and replace it by a sequence of terminal symbol that match the class of words in the input sequence. Thus, in the top down parsing words of the sentence are replaced by their categories replaced by noun phase, verb phase, articles, determiners etc.... These symbols are rewritten as per the grammer rules. Finally the terminal symbols are replaced by language words. This process of parsing keep on continuing until it teach the non terminal symbol.

Example:- Ram went to the park

Production rules:-

  • S--> NP VP
  • NP--> N
  • VP--> V NP
       (OR)

  • S--> NP VP
  • NP--> N/  ARTICLE NP
  • VP--> V / VP PRONOUN
Parsing Algorithum:- To perform the parsing following steps are replaced either till sucessor is obtained or failure is obtained.

  • Select first state from the completeset of possiblities and remove it from the list
  • Genetare the new state by trying every possible options from the related state.
  • Add the state generated in the step to possible list.
Bottom Up Parsing:- In it we start from the non terminal symbols and continuious replacing the sentence with thw sentence with the sentence pharse till reach the root symbol. Bottom-up parsing is reverse parsing of top down parsing.

Example:- Ram went to the park

Production Rules:-

  • NP--> ARTICLE N
  • VP--> N PREPOSITION
  • NP--> VP NP
  • S--> NP VP / N NP  

Monday, 5 September 2016

Rewriting Rules and Production Rules in NLP (Natural Language Processing)

These are the grammer rules which indicates subparts the sentence can be broken. These rules says that certain symbols may be explored in a tree by the sequence of the symbols. Any sentence can be broken according to the production rules.

Examples:- Akhil brought the glass
               S--> NP VP.
               NP--> N / ARTICLE NOUN.
              VP--> V/N ADVERB.

Component of general purpose grammer :-

  1.  Terminal Symbols
  2. Non-Terminals Symbols
  3. Rewriting Rules
Terminal Symbols:- These are the symbols which cant decomposeby the grammer rules.
Non-Terminals Symbols:- Symbols that canbe further broken in to constitutes categories are called non-terminals symbols.
Rewriting Rules:- These are the grammer rewriting rules which describe which word categories can be replaced by which word.

Examples:- Ram went to the park.

Ram-->Noun
Went-->Verb
To-->Preposition
The-->Article
Park-->Noun

S(N(went to the park))
S(N(V(to the park)))
S(N(V(PREPOSION(the park))))
S(N(V(PREPOSION(ARTICLE(park)))))
S(N(V(PREPOSION(ARTICLE(NOUN)))))

Basic terminology of Parsing Tree


There four terminology in parsing tree:-
  1. Label Node
  2. Links
  3. Root Node
  4. Parent Node and Child Nodes.
Label Node:- There are the nodes which represents the basic sentence consiquence categories like Noun phrase, Verb phrase, Noun, verb, Articles etc....
Links:- Nodes are connected by links. Links defines the path between the nodes in the parser tree.It plays the main role in the parser tree.
Root Node:- It is the top node in the parser tree. It corresponds the sentence.
Leaf Nodes:- These are the bottom most nodes in the parser tree. There are many leaf nodes in the parser tree.
Parent and Child Node:- In a link the upper node termed as parent nodeand lower node is called as child node. Hence the same node can be parent node or child node according to the level of parser tree.

Sunday, 4 September 2016

Parsing in NLP (Natural language processing)

Parsing:- The basic terminology to represent the parsing techniques are parsers, symantic analyser and syntatic analyser to extract the meaning of a sentence it needs to be analyse, the analysing of a sentence normally do at two stages

  • Input (sentence)
  • parser 
  • syntactic analyser (structure)
  • semantic analyser (logic)
Syntactic  and Sementic:- These analysers deals with undestaanding and establishing the validity of a sentence according to the basic syntactic rules of a language and semantic analyzer deals with undestanding, meaning, conveyed by the sentence. The tools for performing these analyses are parsers. The parsing techniques are used to generate basic syntactic structure of a sentence and semantic analyser is used to generate syntactic structure are of a sentence.
The parser of NLP can be viewed at two parser in the first place of phase of parser it is taking a instance as a input and then it gives to the semantic analyser. In second phase semantic analyser works on its meaning of words.
The function of a parser is to perform a pre-processing on the input to generate a basic syntactic
structure of natural language sentence. It uses the grammetical information for generating the syntactic structure.
This preprocessing stage is called main stage of parser. Parser perform grouping and labelling of different parts of sentence according to the gammer constituents. It does it in such a way that display their relationship to each other in a useful manners.
parser also checks the validation of a sentence according to the grammer rules. Syntactic analysis does not check the validity according to the sentactic of sentence then if the sentence is syntactically correct ans semantically incorrect, it no need be declared correct after one phase.
Example:- I am going to park.

Steps are:-

  • Language undestanding point and view
  • Sentence structure need to completely analysis
  • Grammer and Parsing techniques
  • Grammer indicate formal specifications
  • A method of analysis sentence according to grammer.

Phases of NLP (Natural Language Processing)

The phases of natural language is broken into various labels such as

  1. Phonological Analysis
  2. Morphology Analysis
  3. Lexical Analysis
  4. Syntatic Analysis
  5. Sementic Analysis
  6. Fragmatic Analysis
Phonological Analysis:- It is the analysis of spoken language. Therefore it deals with speech recognation and generation. The core task of speech recognation and generation system is to take a acoustic waveform as a input and produce output, a string of words. The phonology is the part of natural language analysis which deals with tha phonology. The area computational lingustics that deals with speech analysis is comes under the categories of computational phonology.

Morphology Analysis:- It is the most elementry phase of NLP. It deals with the word formation in this ways individual word is analused according to their components clled morphomes (basic building block of words). We can multiple combination of words using morphomes. In addition non-word taken such as puntuations, special symbols etc.. are seperated from the words. Morphome is the basic grammetical building block that makes words. The study of words structure is reffered to as morphological analysis.
A morphome is defined as a minimal meaningful unit in a language which cant be further broken into small units.
Example:- A word bat is its lowest level. we cannot brek it to  get the meningfull morphome.
                 A boys is the word boy + s . The morphome which form the central part of the word is called stem.In english a word can be made up of one or more morphome.

Lexical Analysis:-  In this phase of NLP, validity of the words according to he lexicon is checked (word is available indictionary or not). Lexicon stands for dictionary. It is the collection of all possible words of the language with their meaning. In NLP the first stage of processing of input is to scan each word in the sentence and compute all relevant lingustic information about that word. The lexicon provides all necessary rules and data for caseing in the first stage of analysis.
The detail of words are noun, verb, adverb, article, singular, plural, helping verbs.

Syntatic Analysis:- Syntax refers to the study of formal relationbetween the words of the sentence. In this phase the validity of the sentence according to the grammer is checked to perform the syntatic analysis the knowledge of grammer and parsing technique is required.  Grammer is the formal specification of rules allowable in the language and parsing is the method of analysis a sentence to determine its structure according to the grammer. Syntatic analysis is done using parsing yechnique a using parsers.

Sementic Analysis:- It deals with the meaning of natural language sentence. In this phase the meaning of the sentence is undestood. If computer wish to communicate by mean of natural laguage, a computatonal representation of these sentence is required to capture the meaning of the sentence.

Fragmatic and Discourse Analysis:-  Fragmatic is the study of relation between the language and context of use. Context of use include such things as identity of people and objects. Fragmatic includes how language is used to refer the people of things or the objects. In this phase od NLP the main intension od speech behind the message is undestood.
NLP offer many ways to refer two entities in a language each form of sentence help n-users how it should be procede with respect to the discourse model and set of belive about the words are replaced with the words. The discourse interpretation requires that one should build an involved represention of discourse state called discourse model. That model contains representation of entities and relationship between entities.


Ambiguity in NLP (Natural Language Processing)

One of the biggest problem in language processing is ambiguity because it refers more than one meaning of the same unit. In the theroy of computational lingustic.
Several types of ambiguity are defined.

  1. Lexical Ambiguity
  2. Syntatic Ambiguity
  3. Sementic Ambiguity
  4. Pragmatic Ambiguity
Lexical Ambiguity:- When one word cn have several different meaning the resulting ambiguity is called lexical or word level ambiguity.To solve this ambiguity additional knowledge regarding to that word is used according to the context.
Example:- 1) Withdraw some money from the bank.
                  2)  Donot go near to the bank.

Syntatic Ambiguity:- Sometimes sentence can be prepared and when parse in more than one way then parser can give multiple meanings of the sentence.
Example:- I saw boy with telescope.

Sematic or Referential Ambiguity:- The use of pronoun and other anaphora can caus reference anaphora . The anaphora are the replacement words that are use on the place of noun in the later part of discourse (full paragraph). We are replacing the noun with pronoun the discourse is the term which indicate multiple coherence sentence.
Example:- Akhil saw a beautuful car in a showroom. He showed it to sai. He bought it.

Pragmatic Ambiguity:- This ambiguity underline the meaning of sentence. This arrives due to different intension of speakon.
Example:- An employee reached to his/her office 9:30 Am but the office tme is 9:00 Am, then boss asked the employee what is the time.

Saturday, 3 September 2016

which database is used by google && Facebook?

 Google uses bigable database in common it also uses MYSQL in minor purpose. Basically the bigtable in not relational database model like MYSQL. It can handle huge hash tables.
Facebook uses
MYSQL:-for storing structured data.
Memcached:-it is memory caching system that speed up the dynamic database drivers websits that caches the objects and helps in fast loading database.Every time it doesn’t goes to the database it just fetches the caching system.
Haystack:-it stores all the pics that are uploaded in a generic object store called haystack.
Cassandra:-it is a apache database.
Scribe;-used for logging system. Varnish:- Facebook uses Varnish to serve photos and profile pictures, handling billions of requests every day. Hiphop:- HipHop for PHP is a set of PHP execution engine.it was created by facebook and make it as opensource in 2010

What Is '1337' Leet

'1337' means "elite", or just "leet" for short. The point of 1337 is to replace all the letters with symbols. It takes a long time to translate between regular English and 1337 even if you know all 26 1337 secrets, unless you're very skilled at it. Here are the 26 1337 secrets.
 A = 4
B = |3
C = (
D = |)
E = 3
 F = |=
G = 6
H = |-|
I = |
J = 9
K = |<
L = 1
M = |v|
N = |/|
(It is backwards because the frontslash under the backspace key is famous for not showing up on many websites.)
O = 0 (zero)
P = |*
Q = 0,
R = |2
S = 5
T = 7
U = |_|
V = |/
W = |/|/
 X = >< (Is normal in words like |-|4x0|2)
Y = `/
Z = 2


NLP (Natural Language Processing) System.

NLP System:-

  1. Inference rule:- It will check the meaning of output generated 
  2. Database handler:- It is used by conversion rules in conversion 
  3. Conversion rules:- Language to language conversion rules
  4. Sematic analyser interpretation:- Rearranging the words according grammer
  5. Parser generation:- Break the speech into words.
Example:- Ram likes to eat mango

  • Ram - Noun
  • likes - Adjective
  • to - Preposition
  • eat - Verb
  • mango - Noun
Role of knowledge in NLP (Natural language processing):- Language processing mainly relays on formal model representation of knowledge related to that language are levels of

  1. Phonology (Phonetics)
  2. Morphology
  3. Syntatic (Syntax)
  4. Pragamatic
  5. Semantic
  6. Discourse
A large number of formal methodlogy of finite state machine formal rules, ethics, logics and probablity theroy is used to capture the knowledge.The main difference between data processing and language processing application is the use of knowledge at the later status. In knowledge regarding various linusics structure can be seperated in the following categories are

  1. Phonology (Phonetics):-  It deals with the sounds of the phonetics (smaller units of a language).
  2. Monphology:- It deals with the knowledge of meaningful components of the words and target meaning in other language.\
  3. Syntatic:- It deals with the knowledge about the structural relationship between the words.
  4. Semantic:- It deals with the knowledge of meaning of the sentence.
  5. Pragmatic:-  It deals with the knowledge of how language is used to accomplish its goals (deals with the paragraph).  
  6. Discourse:- It deals with the knowledge of joining of lingustic unite larges than a single utterance (It deals with the pronouns).



Expert systems in Natural Language Processing

Expert systems are react based on the keywords

  1. ELIZA
  2. LUNAR
  3. SHDRLU
ELISA:- (Counceller) 1960 by weizonbanm
  • User:- Friends are alike
  • ELISA:- In what why
  • User:- They are always begging for something
  • ELISA:- Can you think specific example
  • User:- Well, m friend made me to come here
  • ELISA:- Your friend made you to come here
  • User:-Yes, he says i am unhappy
  • ELISA;- I am sorry to hear that yo are depressed
  • User:- It is true, i am unhappy
SDRLU:- Game playing expert system in (1969-1970)

LUNAR:-Work as the doctor in (1969-1970) Question-Answering System.

Natural language processing (NLP)

It is the field of computer science or artificial intelligence or linguistics concern with the interaction between computer and human.Natural language generation system converts the information from computer database into readiable human language.NLP is the technique that are developed with the aim, the computer can undestand the command given by the human in natural language and perform according to it. 
At present to work with computer the input is required to be given in a formal language. The formal language are those language are thode lnguages which are specially developed to communicate with computer or machine.
Examples:- Forton, c++, pascal etc...
NLP is the sub-field of artificial intelligence and can be applicable on huge number of applications. The general aim if NLP is  undestand our therotical inverstigation of human language: It explore the details of language from computer implementation point of view.
Natural man machine communication:-
  1. User (source language)
  2. Translation system
  3. User (destination language)
Major tasks in NLP are:- 
  • Automatic summerization
  • Foreign language reading
  • Foreign language writing
  • Information extraction
  • Information retrival
  • Name entity recognition
  • Natural language generation
  • Natural language recognition
  • Anaphora recognition
  • Question answering system
  • Spoken dialogue system
  • Text-to-Speech generation system
  • Speech-to-Text generation system
  • Optical characteristics recognition