Scribd
Upload
13 views16 pages

Expert System

The document discusses the concepts of common sense reasoning and expert systems in artificial intelligence. It highlights the importance of both domain-specific and commonsense knowledge for enhancing system performance and outlines the major components of expert systems, including the knowledge base and inference engine. Additionally, it addresses the advantages and limitations of expert systems, emphasizing their specialized focus and challenges in knowledge extraction and maintenance.

Uploaded by

Ayush Rider
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
13 views16 pages

Expert System

The document discusses the concepts of common sense reasoning and expert systems in artificial intelligence. It highlights the importance of both domain-specific and commonsense knowledge for enhancing system performance and outlines the major components of expert systems, including the knowledge base and inference engine. Additionally, it addresses the advantages and limitations of expert systems, emphasizing their specialized focus and challenges in knowledge extraction and maintenance.

Uploaded by

Ayush Rider
Copyright
© All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF or read online on Scribd
216 Anttl ineigence V-Som, Branch) « Response timeisconsisent [Symbolic representation” ig used, Sequential processing isused, Umericrepread Sons, ein Js Distibute (0 Processing | Sed Pg resting an} mH esse and Py tis not good fr Fut tolerant “she ystems, There isa proper explanation] Theres foc any esponse or output that is. itis derived from thee facts and inputs Fr le any explanation, Jane Write a short note on common ser. LER, Des, 2008, une 2014 ay Gina bef nate on common sense reasoning. (RLGRY, Jane ny Ans Computers haveanemrely deserved reputation frlckingeem, A human expert has two main ypes of knowledge ~ () Domainspecitic knowledge (8) Commonsense knowledge Domin-pecifc knowledge deals with ll kinds of knowlege abt Particular domain Al other pices of knowledge that help in reasoning ott than domsin-pecifc knowledge ae termed as common-sense knoe Common-sense knowledge when coupled, with domain-specific , en coupled, with dom: -nowiedge increases the performance capability ofthe system. Tee # ‘number of techniques available that can be used to enable an Al progiat "epesentand reason with common-sense knowledge. As of et howe 2 ‘Program can match the common-sense reasoning powers of a five-year chia To understand the common-sese reasoning consider the follwing life situation. Tommy enjoys shopping in the mall only when the stores" ~~ COMMON SENSE, REASONING, SOME EXaMpie So gE EXAMPLE OF Ene ey a less Toan change in a dynamic environment. And while ones ete 8 is ie ae: eis ot a short notes 7. Write (di) Reasoning. i common (RGRY, June 2016) or 10m sense and reasoning. (RGPY, Nox 2018) se Podele Computer programmers often reuse their code, adapting an old program $tanev suai wth similar structure. Architects dra on ter knowedpe {steely pleasing and useful building ofthe past to design new buildings {1ZtP find pleasing and comfortable. Historians use stories frm the past taf p satesmen, bureaucrats and citizens understand past evens and plan tur, The ability to reason from cases is fundamental to human intelligence. Unit-v 20 {CBR provides a amber of advan ocesses correspond 0 judgements that can have Kew. the wer interface mist often present et Kee acguition an besinpied fc rege ey . a ta mumberofpblens and case hase ramon et tpg ON, pe well a caphcation of the processes by which those the appropriate case. Case-hased approaches to Tea from its experience: (0.29. Hhie a short mote om expert sem (RG 210 Aafia! inotigence (V'Sem. Branch evo of an expert sjstem 10 tht oF & more tational ile. The compiler user cares about ease of input, we reports, but willing 0 tus he compe Dee 20m fiom an expert skill in sch chine dvie or take an intelligent decision about a processing fa Ans, An expert sytem 8 9 et oF program tha may ot 2, dessa he interface show the intemal workings of the Ios eto sl problems ina specialized doin tye Ren SS, ie dem con af the algorithms used, Thin no he in expertise, MF Moma, tet re OE rade reat et gystem, These syst vor the human expertise. - Ms tea eo He iypical expert system, These systems are often not “The British Computer Society's Commitee ofthe y me peen ee Aol for performing 2 task, but rather as a support tool for caper systems as rae the following dfintion "PM pay AE ad a rating process “The embodiment within a a ee ‘nse ane can be considered genealization of exper systems, form thatthe rion ngayon ae represented aaphically in network om, me yh dyeing tough he evo allonng sera onic Cotes ge contol stuctr, They ae mostcommonlyillusaed toa charcterise which many woul gad finden tf the system on demand (0 justify ils own line at eb, eli some game, of solving some puzzle; and th y 380mg ng re? eof payin : and this category M2 ses je bem TPO «ifferent from an expert system. The rules etl intelligible othe inquirer. The style adopted toattantheseskya 3 system iS ‘fre based proeramming Peo apo Two ogi gles bu ater rls or changin ae Knowledge of an expert system is obt Oe ‘coded in a form suitable for the system to 2 sledge acquisition component serves for the construction ofthe src Th pe hon eget ta ee tcp aya, lst vey ena compe oe sources of expertise, teedge acquisition is one of the most complicated parts for the noni op ef heen topcenrvalsincoriaema prota ee per eded to solve problems, expert syster ly 241 Give structure of an expert system, explain its component. torts enayer nen ens inte cis kena ‘ow is «production syste diferent from an expert system 2 DOHAT,. “i8ipciized eld of human endear nace 5 Ans. An expert system, unlike a traditional ‘computer program, is! Mba structure of an ‘81001 that implements a process, but rather it is a representation of at expen s ha essen pert system is shown in fig. $14, The major ha *slemare the knowledge base and the inference engine. Knowledge 220. Artic ineligonce (Sem. 1TBrench) ¥ base consists f predicate ealeulus, facts and rules about the subject at hand. The inference engine consists of all the processes that manipulate the Iinowledge hase to deduce information requested by the user-esolution oF forward or backward chaining, The user interface might consist ‘of some kind of natural longuape Processing system that allows the ser o interact with the system in Fmited form oF natral language The explanation subsystem analyzes the structure ofthe reasoning performed bythe system and explains ito the user, “These four pars ofthe itis used in an application, Jn the consuton of m expe sytem & “Knowledge Engine” es wih an exper inthe fed of appicton in order to reese ti ws norte of the expert in 8 frm tat can be ened tn he Knog base, 0.22. What is an expert system ? Explain all the components of een ‘gsiem, (RGPY, Jane 319 Ans, Expert System — Refer 9 Q29. ‘Components of Expert System — Refer to Q.31 2.38, What are the major components of expert systems ? Explain any ‘ome in deta (RGPN, Dec. 200) Ans. Major Components of Expert Systems ~ Refer o Q31 The Knowledge Base ~ The knowledge base contains facts and mks ahoutsome specialized knowlege domain, An example of asimple know ede base giving family relationships is shown in fig. 5-15. Each fact and le Sentifed witha ame (%. Agr. Fn). Fr ease in reading, te eft 6 in separated from the right bythe implication symbol >. Conjunctson te kt aw given within single parentheses (sublists), and one or more conclsias may follow the implication symbol Variables ae Kentified asa symbul sede ‘by question mark should be noted that eles found in real workings may have many conjunets in the LLH.S. For example, a5 many 08 ei © Phi $14 Base Sactare fay Expert Siem sysem are the pars that consti these Unt-v 22¢ end 2270 une 2) frome 520 int 2) wae) fier 522) Fig. 515 Facts and Rutes in a Simple Knowledge Base InPROLOG, cules are writen as clauses with both a head and body. For ates ll about a patient's symptoms and the corresponding diagnosis “ers might readin English as the rule TF: The patient has a chronie disorder, and the sex ofthe patient is female, and the 880 ofthe patents ess than 30, and ‘be patent shows candition A, and (2B reveals biochemistry condition C. TEN Concudethe patients diagnosis is autoimmune -chronie-hepatti 224 Arta Inteigence (V-Sem., Branch) f Thehisoria ata has te mainine and ie ‘on inference engine has to be kept track off, tain a mn Life eyele of an expert system is shown, ra @. sc Sh a sess ing. S16 rite the characteristics of expert systems, Ans As knowledge intensive programs tf system areas fllows PES Ea tay, nn sation ENR SEM use Knowle athe solution process. “In the knowledge lies the pow exper stem, - " ha daa cont * atom ied yet (i) They should solve dict progr ina tenth human eapes. The i fundamental ican BL no use in having a system that performs like an amateur blu i) They should posses vast anttes of domain om unt-v 225 accept adie, oy. UP and expnd Tee va THe tie bs or eg To aah te Wa Sion oly we ep i, in A use symbolic feesenation for know edge (EP es) ad pion ts nfrece through symbe works OF esemble manipulations of natural language, eto gat closely resemble manipul ag eo metal chen ge, ave © dem actor methods have already been adopted in expert, sa cy a yh om sed ting ss camel systems often reason with melaknowledge. Thats, they Knowledge to the minute deals. For suecessfl operation a i) Exp domain specie know i ede, Thee ae pean gg ih nls abot hmseves, and tsrown owes is and human expert acquit afer long years of professional expen es me ae These priate pees of knowledge Being fred as “hurr ne it i) Tey posses he capac er the nis dei at i incorporated into he system along wit the conventional nose oe a ere pohtocs w ewer pmbienn gstem along ews acti ae viding solutions to the Use’ roblems fom var nue Ch) The mowed s node nd mained set spay fm he cntl rm Aasuch van somplcnenaaes ae prosam isl Ths perme Be ermal sone na Urnemend of he knowledge tse witout esa oh promans. Tis makes posse foe afc oat ae Sine conve popanso pod dfn ype of epen ee fat ‘psc ae knoe etpen em tls soe ey maybe ead Seen tore (0) Tee yo pri ewe of erie each re: A bre seach cigs inact no mame elon eis tench procedure art sdoped Exp sjtemn provi alee inorning tee here Suh proses : (v) Exper sem are eb of epuning hw 3 pet concaon wat achd sn Hy gud infrmton ms comstaice “sro chancleritc hich he Bish Compt Sais" oncxpat cn ave caps ated net dein ~ Trp lier caancethe ee of esse ne human . This far aso ges he wera chance mass and undead sysen oiningaby ny iprovng ewes canes sree pay the role of autor to pupils who would ike to use it san saeco sytem. ‘i A final characteristics fom he point of economists and financial f--They should mint money. Expert systems need heavy investment and [ovsould be considerable ‘Retur-on-Investment’ (ROD, U6 What arte advantage of wing exper sens ? dos Esper pts ae being eloped for ide ary of oma, neva ofexe srems Was wel ealihed by te cy 1980 fee armas of wing exper system ares follows ~ (Expert ssems have inthis Know base vs quits of ‘in pie Knowledge. The knowledge encompases conventional Lore oie om ined mater and exper’. pve aos be Taper stent only serves aan are otis ewe Ba ‘phys the role of knowledge disseminator. (0 atlwsoncto view heen soning pret. The essing totaly tansprent but als provides nse for gusto ie Westiow” This enabies one to look int its interal working, a tis sted the “human window” (@) inthe ofyese years, he knowledge engine aerect “leeds fom the expert, cads it ino a machine understandable om. — 226 Ati! Inetigence (\Som. Banc) Ecteesmtash eam ane i) Oneofaremsocadvanage of wing ESisincrneg ‘which cannot be imagined in case of human ‘experts. sess (ci) takes considerable amount oftime wo tranteran IN ENDET stem vey THs ity ming ng me ste dormin spe who gt Fen ” Orhan specific pens, SM n had moods o Telvan faci. Bary asec He Si 8 fl edged exper. The time for duplication of shor. (i) An expert system plays three major Roles solver attr and an archive. Even though natural lange are very primitive. An ES serves these roles very well, ‘ay not bea good problem solver or & good ttor, (a) Expert systems are highly advantageous inteiisipiny domains where muliple experts are needed, (6) WES. that factors which need more importance a as ven igh priority. In eating with probbilies also the same fotsey afte hhuman exper are adopted, Hence, one can expect a consistent reasoning methodology in an exer ‘stem, Role of probgy Pte ines hile «human by (0.37. Discuss various limitations of expert system. Ars. Preset day expert systems have some limitations, fw ae a fllons- (i) Present day ES focus on very specific tpies only like comput faults, protein cnstallography radiology, diagnostesillsete The majoras being the diel in extracting knowledge, building and maintaining of knowledge bases. (i) Britteness~ Because exper systems only have access tbh specifi domain knowledge. they eannot fall back on more genet nove ‘when the need arses, For example, suppose that we make a mistake in entering dala fr * ‘Radical exper system, and we describe patient who is 130 years old 2 [sieht 40 pounds. Most systems would not be able to guess that ‘have reversed the two fields since the values are not very plausible Unt-v zen sledge can be represented by afew methods only which (on a etek ston mechs hop pages ttiack of Pe jevelopment er wedge ~ Expo ystems ot have ze Ae en-hnotedge~ Expr yt vey inet Gy back Oe about tit Own operation, They tpl cay ie ao nd ions aig ee en gr ewer 8 He ptem Uy See em Tex oa és 1 instal defied oma nagagy oe 2 as fs rson ES ae nS Zio im action of people seh ae psu ative of in expert system in detail and also write its prons and cons, qa Explain eset (RGR, June 2013) Refer t9 Q.29, 0.36 and Q.37. ne Roser pe errs eat ee os eee ee des npc of pe ayn net pale irene prenr aie set teen tetra ae Cnty compa dee edge sneer dpe ete ert cetay eramecretaantae events foe tae eed ‘hs about 150 rules in its. knowledge base with which to diagnose the more seer ens Te sae ae tacettiteanee personnel, Cloning multiple epis for eich o th eight emai biggest benefit is that system cannot rete, and its performance oho mpareamaem a £10 What are the major application areas ofan expr syste? Abs, down tee various applications ‘tm The eve application aren of expe stems se ~ 228. Artic! ntotigence (V:Som, Branch S 6) Analysis (i) Designing (9) Insteetion (8) Planning ox) Rep Applications ofan export system can now be fx business and government, stich as 0) Different types of medical diagnosis, (i) Diagnosis of eompler electronic an ek and clectromechanic alsysen, i) Contot Gv) Diagnosis (i) Monitoring (Prediction ana (ii) Diagnosis of dese electric location systems, (iv) Diagnosis of software development project, (¥) Planning experiments in biology. chemist poet ne ex biology, chemistry, and motu (i) Foreasting crop damage. (si) Henitiation of chemical compound compounds, ictus and cheng (ii) Location of faults in computer and communications sytem (ix) Scheduling ofcustomer orders, job shop production operation, computer resources for operating systems ad various manufacturing tok (3) Evaluation ofan appicans for lending institutions. (i) Assessment of geologic structures from dip meter logs. (Gi) Analysis of structural systems for design or as a reso of earthquake damage, (si) The optimal configuration of components to meet gives specifications fora complex system, (Giv) Estate planning for minimal taxation and oth (28)_Stock and bond potfl (avi) The design of very large scale integration systems, (vii) Numerous military applications ranging from battlefield assessment 1 ocean surveillance GxiiNumerous applications related to space planning and exploration. at, Nero areas ow ining ii casa, poet Jeti, assault and bate and general assistance in nating diet (2) Planning curicula for students Lax) Teaching students specialized tasks like trouble shootin quipment fal, spect gal selection and management. nt -¥ 220 nmriespet ten? Dies i caecercsenapteton i (RGRN, June 2014) fer 029. 0.35 a O40, este ration some of he} eplication of expert sem, on. (RGR, tune 2016) ne reer 0 080 Bs sie 0 are ih ut eons Rn a we HoH pa rhc ofan Exper Spon Rts Ot pptcaan in diferent Aree fon Exe Ste ao. (ult Dicuss expert system tots. sts Expert system tools help in constructing prototypes at rapid speed atcn compromising on the basi characterise, In fat inal expe sytem ware knowledge base 1s empty and the domain expet ims ection withthe knowledge engineer ills up the knowledge ba sally an expert system tol provides the following (i) Powerful editors, menus and graphical inerices for he user fice t communicate {iy The ight envionment for the expr to communist wit it (a) Provides ates for communieaing wh he extemal ernment Gy) Heipsinarapidprotayping (9) Provides sites for ancy of now representation schemes (9a init the expos technology or honing uncon (6 ps in maintaining the system and istrelschemes. 045, Wha are te featres of expert stem tos? ns The main features ofan expt system that one shoul ok in are 8 follows pene (©) oes the too! provide the ight environment or expert tem sey ” : est ap (3) ste toot portable ona varity of hardware platforms ? Gi) Wha is the cos of to! in compaion with others? {v) How frequently do new versions cme and what enanceents bss cach version provide? ipa, Where have these tools been deployed and to what type of Telcos they have been used ? (1) How far the ool is reliable? 4 (ci) What additional resource implementing this? (sii) Will the foo! help in develop time frame ? ea NE 6 be de INeSIEM Within eat (Gx) Does the tool need to be maintained ? O46 Name some succesful exper systems and dns hy re m, ame th various peso expert stems and exp in detail, me (RGRK, Dec. 2004, neem 04 San Mesh nasen Moc oniropc GRY, yy, Mie shor eon MVCN. (REP, sue 00 oy, USER de 2007, a oy Sansome prominent exper stems thei advantage Mar Ne ay Ans. There are many successful expert systems, ’ Some oft ool ete sndtotsbt proved at A aes of eperis Tie ofthe ae — Path forthe development of new techniques systems can be terribly succesful in select (i) DENDRAL does the inferring process of site eleian of chemieal compounds. (i) MYCIN, an expert system for diagnosis of bacterial inftans and effectively handles uncertain data (Gi) XCONRI, a system in us at digital equipment corporation or configuring VAX computer DENDRAL SYSTEM ~ Elucidation of the structure of compounds in organic chemistry was the major task of DENDRAL, an expert stem developed by Stanford University in 1968 MYCIN = MYCIN is an exper system used for diagnosing bacteria infections. The major characteristic of medial domain is the uncertain and precise data coupled wit vast quantities of medical knowledge. MYCIN vas develope at Stanford University to help physicians in denying what bacteria has ben the ease forthe infection and sugest remedial sltions. Control Sructre in Min ~The contol sroctre i 8 goat ected backward chaining of rules. Atay instance Mein resto fix the vale of some clinical parameter. When evaluating a rae fone of the antecedent not known the Myein sets up smal sub-gal to iden the vale, One sale ‘ature is that the ordering ofthe rules is not essential. When the system 230, Arti! ineligence (V-Sem, FFBranch) : Unit-v 23 va ne necessary information, then the Gestion i pt othe ser, i Octet col scare i highly feud enh pe ne aainty Factor Caleutation ~ We discussed Fac Ca al length, how lated in Min. I has nw bee eed 1 ce ang highly sit mee dos nation Facies ~Mycn bs lites awe as (i EPA HOW". WHY" qusion canbe pretense nt sage Maslin Ws be WIT quienes Fane te ACTION part ofthe rule. ten display the promises Eire erowedby herons eb obe HOW eawatere esate ong oes flow We ane as aed a nent of novel Base -Lis Dende nde in ts organ at sto poducon ns One dating xc a auetion rein Mycn isthe cert far sociated wiht i sae Pectin 1SPas el is Eapiheqaet gen “i fact ve sale MYCIN ale neal representation PREMISE : (SAND (SAME CNTXT GRAM GRAMNEG) (SAME CNTXT MORPH ROD) (SAME CNTXT AIR ANAEROBIC) ACTION : (CONCLUDE CNTAT IDENTITY BACTERIODES ‘TALLY.6) English equivalent IF the gram stan ofthe organism is gramneg AND the morphology of the oeanism is od AND the aerobicity ofthe organisms anaerobic THEN there is suggestive evidence (0.6) thatthe organism is bacteriodes ‘Anat fom the prodeton rus, the system asa colton ofits Facts are stored in the form CONTEXT-PARAMETER-VALUE ot [Link] pes ‘The OBJECT also called CONTEXT refers to such things as cultures ‘sen from the patient, the drags administered and so on. These OBJECTS are {uterized by their ATTRIBUTE or PARAMETER andthe value ofthese e228 in VALUE, Associated with each tripe isa certainty factor (CF) Meeh-1and-+1, ACFof+ indicates toil belie, wiles CF of indices veda aN AC of ines 232A iteigence (V-Sem, ranch) E XCON/RI-XCONIR1 is arate tase compute po PAXII7O compte stems Develapety Mebane ea ge sing the ogee language OPS4, the system is currenily being wey et Equipment Corporation, "Ess eBay bass Given customer order, hhave to be made init for te ‘number of diagram showing to be astocated XCON/R determines what fan asons of system functionality how thes various componente Prospector ~Othercrly experts ' ystemprojes ine 4 system thal assists geologists in the discover HS OSPECTOR, (aka XCON), a system used by the Poss, and 247. Write a short note on truth maintenance sytem, @GRY, Dec. 208 cs, Trat maintenance systems (also known a belt eto maintenance systems) are companion components infetenee steno ‘main job of the TMS is to maintain consist mein a tency ofthe knowledge being used by the problem solver and not to perform any inference functions. Ar tsk fees the problem solver and not t perform any inference functions Av ouch, it frees the problem solver from any concerns of consistency an allows ito ‘concentrate on the problem solution aspects.” Fig. 5.17 shows the role played by the TMS as part ofthe problem solver, ‘The inference engine (IE) solves domain problems based on its current belet set, While the TMS maintains the eurenty active belief set. The updating process iS [rae }- incremental. After each inference, | "es information i exchanged between the tv ‘components. The IE tells the TMS what ‘deductions it has made. The TMS, in tum, tasks questions about current beliefs and reasons for failures. It maintains a ‘consistent set of beliefs forthe IE to work with even ifnew knowledge is added and removed. Fig. $.17 Architecture ofthe Problem Solver with a TMS ‘BE. (Eighth Semester) EXAMINATION, June, 2011 (formation Technology Engg, Branch) ARTIFICIAL INTELLIGENCE (Hetive~ I) HT -833 69), any one asin om cach Unit except .[Link] ye: A al aks ay egal Unit (See Unit Page 37,.@.12) or se ate space representation of 8-Purze problem. iM 2) Find ™ (See Unit-l, Page 8,Q.10), spin best first depth Sst and breath fit sarc When would ©) Bale serch be worse than simple breadth Fst search? 10 vit 5:09 Dice asin approaches neg epesation withthe ep examples. (See Unit-tl, Page 72.0.7) 10 Aa propostonl loi? Exit rowlederepeseaon ing a propositional logic, (See U rage 16, Q.12) 10 or 46) Comet te folowing semen oped oie 0 () arts was aman (i Allmen are moral {No moral ves longer han 10 years. ™ (See Unit-H1, Page 80,Q.17) © Ditcate the monotonic nd non-montoni ssn 10 ‘Sec Uae, Pgs 105,08) (a) Bricy describe semantic net. Also explain paritioned semantic net (8) What are scripts ? What are the important components of scripts ? 10 beresaerta 6 (9) Disa 1) ad expan how infomation ean be 18s conceptual dependency and exph YW "presented using conceptual dependency 0 (0) Expin forward and backward reaso ard Feasoning. A peo can proceed either forwatd nr hacheord vor ai chard What factor dee Sn Shove of deton fre parce pe em ae Ute igual Or EMILY, Page 1760.26) (See Unit-ty, » ‘(b) What are the levels of knowledge used in Jangua SEIS. ‘Also rite down the techniques used in NLP MMEDandig 0 vaney SUNN Page 199 9 2% White short notes on any three of the following — (2) Fuzzy logic ‘e () Leaming techniques ©) Common sense (a) Exper system (6) Application of neural networks age 199.03) (See Unit, Page 216.020) (See Unit Page 2186.29) (See Unite, Page 210,021) BE, (Fighth Semester) EXAMINATION, Dewy 2017 nformation Technology Engg. Branch ARTIFICIAL INTELLIGENCE (Elective ar-s33) ‘Note: Attempt one question fiom each Unit, All questions cary equal mats Unie (2) Explain Artificial Inelignce, Also discuss th used (Se Unit. Page (0) Describe how the branch and bound technique could Be used id ‘the shortest solution toa water jug problem, 0 | Page28, 028) areas where i (See Uni or 2. (a) What do you understand by Heuss 2 Explain hil etmibing method. Compare i with gene and test method 10 (See Unit Page 8,06) (Solve the following erypuarthmetic problems 6 a (See Unltd, Page 6.0.78) rious aprroaches of knowledge representation schemes with va pssst (See Unie tl Page 72.0.7) 10 1 cam positional Logie? Explain the knowledge represemation Ys Propo wy wat (Se Unit Page 76.12) 10 Werponesona 1S solution provedur 86.Q24)8 sxplain he oad 4. (9 Earn tion algonthm, Trace the operation ofthe unification (0) Ear rm on cach of the following pairs ofthe Iiteras ia arora) and Ces) acsat, Marcus) oe (See Unit, 1292,0.30) Unit 10 sw Sate and prove Baye’ theorem (See Un Page 11,038 inc are the advantages and disadvantages of conceptual pendcncy? Represent he following sentences inthe eoneept Sependency {0 do ead the ger nthe jungle 1) fam gve the Bock Sta (See UnltAN, Page 138,Q.39) 12 or 6 (Explain semantic net, Give semantic net representation fo the following sentences 12 (0) John's eight i 180 ems (Gi) BIITs eight is 170 ems, (Gi John is taler han Bil. (See Unie (© Wite a short note on fuzzy logie and its operations Page 123,022) 5 Unit (0) Explain Aipha-Beta eu-ofs in game playing with examples. 10 (See Unit-1V, Page 163,013) (© Whatare the components ofa planning system ? Differentiate between ‘ealstate and non-linear planning. (See Unit4V, Page 172,.23) 10 Or * (© Eapain lok word problemi bots with example 10 (See Unle-1V, Page 176,0.26) ‘Aria Itotigence (BE. v (0) Explain natura language processing. Discos any process. (SCC Uni pee PS the voy a 9. (4) What are the var tshngus used in ening 9 " them, (See Unity, a aPin each (Bon Nort Nswn Ditton o (See tnt Page 104) Epin common sme HOME een. (b) Write a short note on expert system. a8 216.046 (See UN Page2 Ig gee ‘ei Semsc ENAMINATIO ‘nformation Techno ARTIFICIAL INTE] N ni ar-8iy June 2012 ranch) ‘CE CEkete—tny Note Attempt any avo parts fom ach Unit. Al questions eary Unit 4+ (2) Youate sven 2jupsa4 gallon uganda 3 allan one Neher ag, Imesutig mater on Theresa pump that can be wed ah AME With water. How can you get exactly 2 sallon jug? Write the production rte for ‘a ks Hn oF wae the heabove problem ig (See Unit Page 28.92 (©) Explain the purpose of production system with examples, Aso coving the various types of production system, (See Unit Page 28.24 10 (©) Whatiscontol strategy ? What ane the characterises ofa good cont see (Sect nil Pages. e410 Us 2 (4) Soe purse problem using Best first searh algorithm; 10 Te) o 21513] + FepsTo tele] [ats (See Unit} Page 1.042) (0) What do you understand bythe representation of knowledge ? Exp the varus problems in epresening now ede 0 uaweuen (Ser ist rage 7.010 ©) Convert the following sentences to predicate logic y (9 AllRomans wer iter loyal t Caesar or hte hin. {Sec Unit Page 82,2 (0 (Everyone is loyal to someone, Se Uni Page 42,021 ‘Aric tegonce (BE. vsSem) ss tied 10 woe Hsee Unie Page a2, 0.219) ar was rer 1USee Unit Page 2.021) wens Uniti declarative a procera frames, 10 a esi (See Unit, Page 131,038) wal dependeney and explain how infurmation can be 1 coneeptul dependency ? it (See Unit, Page 132,036) ww Fuzzy logic is beneficial over classical probability py Dacess ted repre cp lain theo (See Unit. Page 144,0.55) 10 Unitary sine vinimax strategy of game playing techniques with example 1.) Bsn 10 ty Fspan black word problem in robotics with example. 10 Unity 4 (0) Explin Expert system in detail and also write its prons and cons. 10 (See Unit-V, Page 227.38) (t) Explain explanation based learning and learning from analogy with cxample. (See Univ, Page 202.Q.13) 0 (6) Whats neural network ? Writ its various applications 10 (See Unite, Page 213,0.22) ‘BE, (Eighth Semester) EXAMINATION, June, 2013 (formation Technology Engg. Braneh) ARTIFICIAL INTELLIGENCE (Bective 11D (7-833) \*€=() Anempt one question fom cach unt (8) Please do not do any rough work on questions paper Unitet types of production systom & giveits characteristics. meee (See Unie, Page 23, Q.24) 10 (© Find the OPEN and CLOSED lst odes forthe diagram given below ~ o 1) Det Aric Ittinone (BE. V-Som) / @ (eeu or Page a39, 2. (a) Suppose that we have got to jigs having capacities og liges prove that using them We can measure Le 353 using breadth fist search tse (©) Explsin AO algorithm, lives a3 Se ands tne 238.039 19 BEES. Q 19 Uni 3. (a) Comer te following sentences to predicate logic (6) Marcus was a man (@) Allpompeians were Romans (No moral lives longer than 150 years (iv) isan even number (0) All even numbers are divisible by 2. (See Unit, Pages, (©) Explain monotonic and non-monotonic reasoning with exanploe (Se Unittl Page 105,953) or 4: (@) Explain knowledge representation & problems associated witht with examples, GeeUnied (©) Convert the fllowing 1 clause form PCY Oy. 3). . Unit 5 (2) Give dhe concepul dependency representation of Ram gavea pets yam. (See Unie-, Page 139,04) 10 () Explin fuzzy logic with examples. (See Unit-ll, Page 142, Q.8)10 Or (a) Represeatthe following sentences in the form of Semantic Nebwo 10 (}) Mobun Bsgan & East Bengal played football with each ote. (@) East Bengal defeated Mohn Bagan by 2 goals oI (©) Explain forward & backward reasoning with examples. uy (See Unie I, Page 150,057 (See Unie, 7a) ‘Write the block world descr ° ieiervesiseres (Block Bison block A = ‘Artic! iotigence (B., V-Sem) aderstoading with example, (See Unit, Page 180, ane meaning of the following in game playing 2. (a) Explain aici intigence Als discus th ary % vied 3 hy (See Unicn peg (©) Compare breadth fist search and depth puoi“ Page the SAAC tech OS Unittt OEM Paget) 3 (@) Esplin the various approaches of knowledge rpg, y with the help of examples. ann (Die beeen monsonic and nea Pa on (See Unica rageng? + 0 Bile ows UPS td agen, (0) Writ short note— SeUMICHL Pages 9 Reson (See Units Paet6 2) ,Page99. a9 (i) inerencng voit (3) Sate and prove Bayes cre. (SeeUnist, Pager (©) Explain semantic net Give semantic net pretense ee following sentences 7 () Rams eight is 165 xm (Sees tls han Geta. (See Unie, ge 12,020 or 5. (a) Explain fury logic and ts operations, (Se Unit- I, Page 48,057 () Explain the concept of forward and backward reasoning with be help of example, (See Unie, Page 150,057 Unierv (3) Give a brie note on natural language processing. Also dics application arse. (See Unit, Page 187.0397 (®) Esplin minimax procedure ingame playing with cxamples. | 7 (See Unity Page 1580) or (Palin eck word pobemin oot with xml 7 See Unit Page 176 (Diu componens of planing ton?! (Gee Unit, Page 17202 Uniev (0) Witiskeing? Explain varios techniques being sed ine” “Sco Unit VPage 201, 2 eS ‘Ariiil intligance (8... V-Sem) acteristic and applications.7 » Discus its ea tis export patent? (See Unit¥, Page 229,Q-41) or stand by Neural Network ? Discuss the mW jou unders 0. 0) Whe oon af eral networks. . 1 application (See Univ, Page 213, 0.22) 7 on sense reasoning, brief note on comme ive (See Unl-¥; Page 216, 0.26) o 'BE. (Eighth Semester) EXAMINATION, June, 2015 (information Technology Engg, Branch) ARTIFICIAL INTELLIGENCE (Elective IT) (ar -833) Aten one quesion fom ach un. Each unit have intemal choice ‘Scmedativ require. nie _ tm A* does nt tmite unl «gol node is seleted for ‘Zpumon. Hevea peth tothe goal nde might beached long bee a ode sete for expan. Why doesnot teminte ison at ol noe as been fund? aa your ans ih snemnple (See Unie, Pages Q55)7 0) Gian example of problem far which bread fist search Would ‘ork eter ha depts sere, Give an example of problem for ‘ch depint serch woud worker han east earch, (See Une Page22,0.3)7 or 2. (a) Write AO* algorithm, (See Unit, PageS9, 0.68) 7 (©) Describe the behaviour ofa revised version ofthe steepest ascent billctimbing algorithm in which step 2 () i replaced by “set current sate to best successor (Gee Unie, Page62,Q.71)7 Unie 3) Consider the following sentences. 7 (Joh tikes al kinds of food, (i) Apples are food (Gi) Chicken is food, (@) Any thing any one eats and isnot killed by is food. Translate ‘hese sentences into formulas in predicate logic (See Unie, Page 81, Q20) (© Wt comparison between monotonic and non-monotonic reasoning 7 (See Unittl, Page » c age 105,0.52) ‘Artal bligance (BE. V-Som) ya 4. (2) Draw tthe ar show ht the fllowing ia vay seri Noman wes of neural NEN itis humid then it will rain and since itis humid 4 ) ive te 8 ¢See Unt ‘4 har o winston’s Fearing program ee (0) rove the lowing orem ing dette ta spin abot Mee UN Page 198.0 O) roma Bac SS on eee tow i) From An B.A >: pent sytem rent fom aration roe? Ho Se eee seo Heese zc 1m a © ("Sachin pl fru ndansand asta sang sro sea {Sec Une-v Page 218,020) fone one iy od reson = 1 case-based reasoning. - (3) Stet is facut member. yeas ao (See Un¥,Page217,028) (0) Assume universe of discourse foF set oF salary struc oF k) per month i defined as HEP SUC i tag, Us (2,5, 10,20, 30,40, 50, 60, 70, $0) 7 The fuzy sets defined on U based on salavis are Poor = {@, ($1), (10, 08), (20, 05). (30, 0.29} Average = 110,02), 20,05), 8, 0.6, 40, 08), (50,09, (Eighth Semester) EXAMINATION, June 2016 {nformation Technology Engg. Branch) ARTIFICIAL INTELLIGENCE (Elective) (aT -833) BE. Rich = (40, 02), (50, 0.5, (60, 0.8), 70,1), (80, 9 “22 ( Answer five questions, In each question part A, B, C is Daw the graph of al tes fizzy sts on the sone coi aes Srapulsory and D part has internal choice Find the membership ale of your age, om (ay Altpats ofeach questions are to be attempted at one place. or Ai) allquesions cany equal marks, ot of sehich part A and 8 6 @) Comic semantic net representations for he following 7 (Max, 50 words) carry 2 marks, part C (Max, 100 words) carry (6) Pompian (aru), Blacksmith (marcus) 2 marks, part D (Max. 400 words) carry 7 marks (@) Mary eve the ercen flowered vse to he fa (6 Exzapt numerical, Derivation, Design and Drawing ete . (See Uni (0) Lista few of the ask domain of AL (See Unit, Page s,Q.) (0) Wiite Bayes thorem ao discus isu (0) Name various types of control strategies. (See Unit}, Page 67, 0-76) (SeeUniitt,Pageit,Q3)_() Diseussbenefits and short comings of Breath fis search techniques unity (See Unit, Page 30, Taree | nit, Page30,Q31) (Hoy wold th inna peer tveto be mde te usd Etna clas “Water Jug Problem Dene easy ae by 8 pogom paying the or fur neon gan rar an for this problem (ee Unit, Page 18, 18) one 7 or (©) When your ind elim ty , Explain th When sour end cain tat LSI produces the best approximaon sin the meaning ofa production system in anificial intelligence te original tm document mars. You woul ie 02) Wihstamol; Whatarthernain eomponen ors production system So make he statement meaning. What are these? 7 (0) White any ieee ne See Uni Page 25, 0.26) Coe ‘me Wits any evo ditference between procedural and declaesi soli pha lee realitalphe-bta search procedure See Unit V, Page 163,011)7 (© Whats resoltion princi (Gee Unit, Page 69,Q.3 0) laenity toe ee te“ee Bae. (©) Explain non-monotonie cecseni (Sce Unit-M1, Page 86, 0.24 tty wo ways in whi ir wnotonic reasoning, (See Uni-ll, Page Demet MLPysams can make ie eas () Explain ana 1. Page 103, Q.48 ‘write unification algorithm, (See Unit-Il, Page 91,Q 26 ‘ or ay test eine (BE, Sem) Consider th owing sentences {0 "John Mesa Kinds 0 food (a) Apes ae fod {ih Anhinga et ands ot Ki {i Chicken fo “Yi fg 3. (a) Write any heap (©) Explain forward reasoning. (@) What se frame ? How (See Uni me on Sma ney 52, (0 Bin concn dependiny ? 0 NEM ag, ‘ston oh eee, Jobn age eee ee ely re Construct a semantic nets o (0) The dos bit the mail ease (i) Every dog has bitten every mai ‘What is alpha-t i ©) Mishletmameng? Sumer eet ao ‘s important in A. (See UnitclV. Pane eet (© Banani each oct len eee HO (See Unie 1 Page 1 on Page 6, ; tae ee oy Remieety, 7 PMN nog oesing Lit ia (3) Menton sone (See Unity, Page eee aptamer (ertatPge2. 00 (See Unit-V, Page 199,09) network ©) List atious © Giveany eign, re aplition ef acral (© Explain eb (See Unity Page210,08 i te bse ae cenit. Page 218,020 ri eigen of iene Se hue examples (Gee Unity, Page29,049 Write ashornous OF es © Common sense @) Resning (iy (ee Une Page 217,027 SXAMINATION: eet ota Eas Beane) ve apie INTELLIGENCE Ele mm) eile aT -833), sc question eases caus marks sive questions: be data ym solving echnigues (i sme a Seon ON pages) oe gues. haracteristics. cee ive its types and ch ste AB, Page 28,0.20)7 See Unit ee sion usin of stony 2 i ive este Tae See Unit Page 14 Q15)7 eels problem en il Clibing sport How 6 from Hill pest lining ago. How different Wrest co Gee Unit-t, Page 42,Q.51)7 ((Sce Unit-II, Page 72, Q i yao aha oot Sah ert) Sac rae re moe : (Every dog bites a mail carrier {@) John gave a book to many o (seus age 19,016 Describe Baye’s theorem. Show how it ‘ Sa es Sse ante cra 1 Sekine ep on ingagunteintog he) () Describe some of the reset a37) procedure, ‘sfsement ec iques used in minimax search (is) SeeUniey, Page 169,9.18)7 Ari teigence (8. V-Sen) (0) Explain the cone oF amine by ston wing (See Unig pate 8, Write soa aos on Pa (2) AO* Seach (srt , (0) Fuze operations (Seen iey (6) SiperedLeaming Soe ae s (SeeUne nag ei 5 @) (Cas BE. V Someate? EXAMINATIONS Choice Rased Grading ‘ARTIFICIAL INTE ‘iss 015 LiGENCE Note: i) Anompt any five questions i) All questions cary equal mack, 1. a) Dios the characterises of production sytem, (Set aan rhea (eta ‘algorithm with its merits and demerits Tae: idered to hemor pce than the tadonal search techniques ? (See Unie. age egg (2) What are various approaches to knowledge repretentager (See Unie Pape. 95 lausal form Show tat le using resolton — (©) Expisin your own criteria for compute incthgnt. 2 (a) Wrtedown the at (See Unit (©) Why heuristic search techniques are consi (©) Translate the following statements in el Predicate suppor (notebook. cup) ist (i) IExis on the top of. suppor ( Ife above y and thy are uching each eh son pt {GA cap isabove a notebook " 69) cupis touching @ notebook, (seetne 4) Siseand pve Bases ese charm, ce %) Whstisstpc? Wie eset of cine Sh s 5) Explain the Charnes Semantic nets and make the partioned seman the followingsenence = ery dog into as bien tease (©) Show 3 concepn (Gee Unie Page 8.017 senonce *PH dependency representation ofthe fll “lh prevent 19h Prevent Mary from giving a book to bill” (See Unit n,pagets.0e 4) EEE" srnaence 08. 050) a in with the help of an 1 xp ie Te aD) ec Unit hn a ax al ‘ > Jos sem write anning S¥° ge 172, Q.22) 10 ot capnsto HAN a age 73.0 oy Wh ing. Discuss its appli “pee Unit-V. Page 205, 0-14) table example. nite, Page 186,034) aso eat plain exlaaion = ‘1 rage processing lina 9" ow B : ny te) ee ee yt tt 5 7 coeur 10 ©) MvCl i (soe Unite1T, Page 128,931) {See Unit-V.Page210.0.21) (ee Unit. I, Page +4, 0.55) Gy aot algoeit Frames © of ne (a Application of ilstinbing ural network © Coamis © CHES) Bary Semester AMINATION, Nov. 2018 oie Bane Grading System (CBGS) iGeMCIAL INTELLIGENCE Sues (Aton any ve estos ti) Allqustions cary equal marks Gi) Assamese value for missing dat, i ny (0) Bese various task of Al Wht types of characteristics should ald the good production system? (See Unie, Page 25, Q.25) (b) Discuss the block werld problem with their production rue (8) ow do we minimize the tt esimated solution os using A search? Explain, (See Unit-l, Page 54, Q.65) (© Explain heuristic fonction and its advantages for optimal search, ote (Gee Unit, Page 40, Q.44) ‘knowledge. Explain the knowledge representation (See Unit-1, Page 74, Q.8) feaon auth or predict loge and ee eo Ssekward chaining and Forward chaining, 4 Mifereniae (SeeUnie thi Predicate logic for th ne kind ot Gi) Apples are food, oF foa (iy hess el (Amnon a edo: by a5) techniques in AL % 2091, 1515 029,59 1¢ following — maeeel is food, Atif al Intligence (B.E., V-Sem) (v) Bill cats peanut and still alive, (Vi) Sue eats everything Bil] Cats, (b) What is refutation 9 p, TOVe that Joh chaining, (a) Explain Partition se het for the following Teasoning (i) Emu are birds Gi) Typically bina Mantic n, Is fly and have Wings (i) Emus Tun, Up, . 5 Pen. yy (See Unity ny, Pals Ling letWork, and rep See y, 0 Hike nity y, : ae © Unit Pager, : () Explain classes ang object in frame ang Bive advan, disadvantages of frame, ¢Unitan, Page = (a) Construct the Conceptual dependency "eDresenaay a following - (i) John gives Mary a book. (i) Ram Bave flower to Sita, (See Unien, Page 135 9, (b) Mention the frame Manipulation Primitives, Define forward and backward chaining, Differentiate the same See Unit, Page sy, 08 (a) Name Some prominent &xpert system, Sive their advantages, (See Unit-V, Page 230,94 () Explain min max search Procedure. Discuss the applications of 7 SIWork in medical fe (See Unit-V, V; Page 161,213; Ql. Explain the following — 2 © Common sense Amdreasoning (gag Unit-y, eum (ii Backtracking (See Unit-, Page 68,0! Ci) Application of natural language Processing in (See Unit-IV, Page 187, () Fuzzy logic Proposition and Membership function 05 (See Unit-IM, Page 149 (16) tt

You might also like