Intelligent Information Systems - uni-bonn.de© 2016 Prof. Dr. Rainer Manthey Intelligent...

©© 2016 Prof. Dr. Rainer Manthey2016 Prof. Dr. Rainer Manthey Intelligent Information SystemsIntelligent Information Systems 1111

IIS WS 2016/17IIS WS 2016/17

Prof. Dr. Rainer MantheyProf. Dr. Rainer Manthey

Intelligent Information Systems

(MA-INF 3203)

Intelligent Information SystemsIntelligent Information Systems

(MA(MA --INF 3203)INF 3203)

WS 2016/17 WS 2016/17


Vita Rainer MantheyVita Rainer Manthey

1953 Wilhelmshaven1953 19531953 WilhelmshavenWilhelmshavenWilhelmshaven

1973 Kiel1973 Kiel1973 Kiel1973 Kiel

1984 München

1992 Bonn1992 Bonn1992 Bonn1992 Bonn

University of KielUniversity of KielInformatics/MathematicsInformatics/Mathematics

SStudent (Diploma 1979)tudent (Diploma 1979)RResearch assistant (PhD 1984)esearch assistant (PhD 1984)

European ComputerEuropean Computer--Industry Industry Research Centre (ECRC)Research Centre (ECRC)

RResearcher/esearcher/TTeamleadereamleader

University of BonnUniversity of BonnPProfessorrofessor

1984 M1984 Müünchennchen


Modules Offered by the IDB Group Modules Offered by the IDB Group

Temporal Information

Systems(MA-INF 3302)

Temporal Temporal Information Information

SystemsSystems(MA(MA --INF 3302)INF 3302)

Seminar

Selected Topics inIntelligent IS

(MA-INF 3210)

SeminarSeminar

Selected Topics inSelected Topics inIntelligent ISIntelligent IS

(MA(MA --INF 3210)INF 3210)

Intelligent Information

Systems(MA-INF 3203)

Intelligent Intelligent Information Information

SystemsSystems(MA(MA --INF 3203)INF 3203)

Lab

Intelligent Information Systems

(MA-INF 3313)

LabLab

Intelligent Information Intelligent Information SystemsSystems

(MA(MA --INF 3313)INF 3313)

WS

SS

WS+SS

IDB (Intelligent Databases) GroupIDB (Intelligent Databases) Group::Prof. Dr. Rainer Manthey Prof. Dr. Rainer Manthey PD Dr. Andreas BehrendPD Dr. Andreas BehrendSahar Vahdati, MScSahar Vahdati, MSc**

(* partially with Prof. Auer(* partially with Prof. Auer‘‘ s group)s group)

NEWNEW : 4 hrs lecture/week: 4 hrs lecture/week

Scientific Writing

Scientific Scientific WritingWriting

NEW: 2 hrs lecture/week (?)NEW: 2 hrs lecture/week (?)

IDB: IDB: „„ Remains of the DayRemains of the Day““


Prof. Dr. Rainer Manthey:Prof. Dr. Rainer Manthey:

PD Dr. Andreas Behrend:PD Dr. Andreas Behrend:

•• 28.2.201928.2.2019: : Day of Day of RRetirementetirement•• => 5 semesters of teaching left (only in MSc program):=> 5 semesters of teaching left (only in MSc program):

WSWS 16/17, 16/17, SSSS17, 17, WSWS 17/18, 17/18, SSSS18, 18, WSWS 18/1918/19•• => Supervision of master thesises=> Supervision of master thesisesends 30.9.2018ends 30.9.2018


(German: (German: PPrivatrivatddozent)ozent)

•• PD/Habilitation: Full qualifications for any kind ofPD/Habilitation: Full qualifications for any kind ofacademic teaching (incl. thesis supervision), indepenacademic teaching (incl. thesis supervision), indepen--dent teaching scheduledent teaching schedule

•• Position at Uni Bonn Position at Uni Bonn endsendsat at 30.6.201830.6.2018(at latest)(at latest)•• At most 3 semesters of teaching left:At most 3 semesters of teaching left:WSWS 16/17, 16/17, SSSS17, 17, WSWS 17/1817/18•• Last two years: Guest professor at universities in Dresden, MLast two years: Guest professor at universities in Dresden, Marburg, Hallearburg, Halle•• TThis semester: teaching in Bachelor and MSc LSIhis semester: teaching in Bachelor and MSc LSI•• Next semester(s): teaching still openNext semester(s): teaching still open


Department I: Department II: Department I: Department II: Department III:Department III:

Prof. Ro. KleinProf. Ro. Klein Prof. Re. Klein Prof. WrobelProf. Re. Klein Prof. WrobelProf. KratschProf. Kratsch Prof. Weber Prof. Weber Prof. MantheyProf. Manthey

Prof. Hullin Prof. GallProf. Hullin Prof. GallJun.Jun.--Prof. Schultz Prof. Schultz Prof. AuerProf. AuerJun.Jun.--Prof. YaoProf. Yao Prof. LehmannProf. Lehmann

Department IV: Department V: Department VI:Department IV: Department V: Department VI:

Prof. Martini Prof. RProf. Martini Prof. Rööglin glin Prof. BehnkeProf. BehnkeProf. MeierProf. Meier Prof. BlumProf. Blum Prof. AnlaufProf. AnlaufProf. SmithProf. Smith Prof. BennewitzProf. Bennewitzapl. Prof. Kurthapl. Prof. KurthJun.Jun.--Prof. ReinhardtProf. Reinhardt

((Department IIIDepartment III: in R: in Röömerstramerstraßße; all e; all other departmentsother departments: in Friedrich: in Friedrich--EbertEbert--Allee)Allee)

““ GeographyGeography”” of Our Instituteof Our Institute

Groups in Department III

•• Main area of Research (and Teaching) in Department III: Main area of Research (and Teaching) in Department III: Intelligent SystemsIntelligent Systems

•• Until 2013: Information Systems and Software EngineeringUntil 2013: Information Systems and Software Engineering

•• Since then: Since then: New department headNew department headProf. Wrobel Prof. Wrobel •• Who is also Director of the Fraunhofer Institute IAIS and DireWho is also Director of the Fraunhofer Institute IAIS and Director of Bctor of B--ITIT

((IAISIAIS: Intelligent Analysis and Information Systems;: Intelligent Analysis and Information Systems;BB--ITIT: Bonn: Bonn--Aachen International Center for Information Technology)Aachen International Center for Information Technology)

•• Recently: Two big Recently: Two big new research groupsnew research groupsfounded (associated with IAIS as well)founded (associated with IAIS as well)•• EIS EIS –– Enterprise Information SystemsEnterprise Information Systems(Prof. Auer) (Prof. Auer) ~ 40 members~ 40 members•• SDA SDA –– Smart Data AnalyticsSmart Data Analytics(Prof. Lehmann) (Prof. Lehmann) ~ 20 members~ 20 members

•• Since 2013: Since 2013: Computer VisionComputer Visiongroup (Prof. Gall) group (Prof. Gall) 10 members10 members•• Since 1992: Since 1992: Intelligent DBIntelligent DBgroup (Prof. Manthey) group (Prof. Manthey) 2 members2 members

•• Teaching (main lectures in MSc): Teaching (main lectures in MSc): WrobelWrobel: Machine Learning, Data Science & Big Data: Machine Learning, Data Science & Big DataAuerAuer: Semantic DataWeb Technologies, Enterprise Information Systems: Semantic DataWeb Technologies, Enterprise Information SystemsLehmannLehmann:: Knowledge Graph AnalysisKnowledge Graph AnalysisGallGall: Computer Vision I + II: Computer Vision I + IIMantheyManthey: Intelligent IS, Temporal IS: Intelligent IS, Temporal IS



OrganisationOrganisation

OrganisationOrganisationOrganisation

Intelligent Information SystemsIntelligent Information SystemsIntelligent Information Systems

WS 2016/17WS 2016/17

•• Every week: 4 hours lecture + 2 hours exercises (9 credits)Every week: 4 hours lecture + 2 hours exercises (9 credits)

•• LecturesLectures: : Monday Monday afternoon (14:15 afternoon (14:15 –– 15:45) + 15:45) + Wednesday Wednesday morning (10:morning (10:3030 –– 12:00)12:00)

•• ExercisesExercises: : Wednesday Wednesday afternoon (afternoon (12:4512:45–– 14:1514:15))

•• Lecture and exercises in the Lecture and exercises in the samesameroomroom (A 207).(A 207).

•• On On WednesdaysWednesdaysspecial special timingtiming conventions:conventions:•• Lecture starts 15 minutes Lecture starts 15 minutes later later than usual: than usual:

10:3010:30rather than 10:15 a.m..rather than 10:15 a.m..•• 45 minutes break 45 minutes break between lecture and exercises between lecture and exercises

(rather than just 30 minutes)(rather than just 30 minutes)•• Exercises start 30 minutes Exercises start 30 minutes later later than usual: than usual:

12:12:4545 rather than 12:15.rather than 12:15.•• Exercises end 30 minutes Exercises end 30 minutes later later than usual: than usual:

14:1514:15rather than 13:45.rather than 13:45.•• NextNext lecture (Prof. Auer) starts 14:lecture (Prof. Auer) starts 14:3030 rather than 14:15!rather than 14:15!


Weekly ScheduleWeekly Schedule

Intelligent Information SystemsIntelligent Information Systems 99©© 2015 Prof. Dr. Rainer Manthey2015 Prof. Dr. Rainer Manthey 99

Semester Schedule (WS 2016/17)Semester Schedule (WS 2016/17)

MondayMonday WednesdayWednesday

OctoberOctober 1717 1919

2424 2626

NovemberNovember 3131 22

77 99

1414 1616

2121 2323

2828 3030

DecemberDecember 55 77

1212 1414

1919 2121

JanuaryJanuary 1111 1313

1818 2020

2525 2727

FebruaryFebruary 11 33

88 10 10

18 lectures18 lectures8 exercises8 exercises

Dies academicusDies academicus

Begin of exercisesBegin of exercises

10 lectures10 lectures5 exercises5 exercises

Xmas break (2 weeks)Xmas break (2 weeks)

©© 2016 Prof. Dr. Rainer Manthey2016 Prof. Dr. Rainer Manthey

20162016

20172017


Exercises and Exams: Exercises and Exams: „„ Rules of the GameRules of the Game““

•• ExercisesExercises::•• In the In the samesameroom every Wednesday, following the lecture after 45 minutes brroom every Wednesday, following the lecture after 45 minutes break,eak,

for the entire auditorium, for the entire auditorium, nono small groups. small groups. •• Exercises held by Prof. Manthey and/or Exercises held by Prof. Manthey and/or Sahar Vahdati.Sahar Vahdati.•• Goals: Goals:

•• To make you fit for the exam!To make you fit for the exam!•• Attention!Attention! There are There are too fewtoo fewexercise weeks (for a long lecture)!exercise weeks (for a long lecture)!•• To provide some To provide some „„ hands onhands on““ experience with theoretically introduced concepts.experience with theoretically introduced concepts.

•• ParticipationParticipationwill will notnot be checked, but is strongly be checked, but is strongly recommendedrecommended!!!!•• NoNo prerequisites for getting prerequisites for getting admissionadmissionto exams! to exams! •• NoNo „„ homeworkhomework““ to be delivered, but motivation/encouragement for individualto be delivered, but motivation/encouragement for individual

activity provided in exercises. activity provided in exercises. •• NoNo individual feedback possible.individual feedback possible.

•• ExamsExams::•• RegistrationRegistrationfor exams: December 1 till 22 for exams: December 1 till 22

(online via BASIS)(online via BASIS)•• Written examsWritten examsfor both exam dates for both exam dates

(120 minutes duration)(120 minutes duration)•• Exam Exam dates to be determineddates to be determined: :

Most likely end of February + end of MarchMost likely end of February + end of March


IIS HomepageIIS Homepage

https://pages.iai.uni-bonn.de/manthey_rainer/IIS_1617/https://pages.iai.unihttps://pages.iai.uni--bonn.de/manthey_rainer/IIS_1617/bonn.de/manthey_rainer/IIS_1617/

Slides for download!Slides for download!Slides for download!


No Book, just Slides!No Book, just Slides!

There is There is no textbookno textbookwhich could be recommended for this lecture . . .which could be recommended for this lecture . . .

. . . . . . just the slidesjust the slidesserve as a substitute insteadserve as a substitute instead(representing a compromise between a good(representing a compromise between a goodbackground presentation and too much text)background presentation and too much text)


A Word of WarningA Word of Warning

But:

Only a small fraction of the attendees will have achance to get a place inseminars or to do a masterthesis in this area!

But:But:

Only a Only a small fractionsmall fractionof of the attendees will have athe attendees will have achance to get a place inchance to get a place inseminars or to do a masterseminars or to do a masterthesis thesis in this areain this area!!

IIS 2015: Over 100 participants in the exam !!IIS 2015: Over IIS 2015: Over 100 100 participants in the exam !!participants in the exam !!


BackgroundBackground

BackgroundBackgroundBackground


WS 2016/17WS 2016/17


Information Systems: The DBInformation Systems: The DB--centered Viewcentered View

Information SystemInformation SystemInformation System

Database SystemDatabase SystemDatabase System

++ External media External media of communicationof communication

++ ApplicationApplication --specificspecificmethodsmethods

This is the most commonly agreed view on the concept of an IS in informatics –provided people agree on the meaning of DBS!!

This is the most commonly agreed view on the concept of an IS inThis is the most commonly agreed view on the concept of an IS ininformatics informatics ––provided people agree on the meaning of DBS!!provided people agree on the meaning of DBS!!


Databases and Database SystemsDatabases and Database Systems

DBDBDBMSDBMS

Database SystemDatabase SystemDatabase System

. . . .

Users and Users and application programmesapplication programmes

DBMS: DBMS: Data Base Management System

Database(s)Database(s)

(Many powerful (Many powerful applicationapplication--independentindependentservices: schema mgt, query optimization,services: schema mgt, query optimization,storage mgt, transaction mgt, etc.)storage mgt, transaction mgt, etc.)


IDBS rather than IISIDBS rather than IIS

This lecture will be more accurately concerned withThis lecture will be more accurately concerned with

rather than withrather than with

The naming of the module is more a matter of convention rather tThe naming of the module is more a matter of convention rather than precision!han precision!

Intelligent Intelligent DatabaseDatabaseSystemsSystems

Intelligent Information SystemsIntelligent Information Systems


Query Languages vs. Programming LanguagesQuery Languages vs. Programming Languages

DBDBDBMSDBMS

Data DictionaryData DictionaryData Dictionary

•• „„ RealReal““ DBMS support a separate kind of DBDBMS support a separate kind of DB--specific specific „„ programming languageprogramming language““ forforaccessing and manipulating data in the DB: accessing and manipulating data in the DB: query languagequery language

•• In contrast to the external imperative programming languages,In contrast to the external imperative programming languages,a query language isa query language isusually a usually a declarativedeclarativelanguage, the performance of which is optimised by the DBMS.language, the performance of which is optimised by the DBMS.

•• „„ ProgramsPrograms““ of the query language may be stored in the of the query language may be stored in the data dictionarydata dictionarywithin the DB. within the DB.

ImperativeImperativeprogrammingprogramming

languagelanguage

DeclarativeDeclarativequeryquery

languagelanguage

InterpreterInterpreter


Relational Data Model and SQLRelational Data Model and SQL

•• The most widely used data model nowadaysThe most widely used data model nowadaysis the is the relational modelrelational model(introduced around 1970).(introduced around 1970).Relations are the mathematical basis for dataRelations are the mathematical basis for datarepresented in tables (rows/columns).represented in tables (rows/columns).

•• All relational DBMS support a predominantAll relational DBMS support a predominantdeclarative query language based on declarative query language based on logicallogical andandalgebraicalgebraicoperators:operators:

SQLSQL (Structured Query Language)(Structured Query Language)


Background in Relational Databases and SQL: Strictly Necessary !Background in Relational Databases and SQL: Strictly Necessary !

A A good backgroundgood backgroundin relational databases andin relational databases andin SQL is expected from everybody attendingin SQL is expected from everybody attendingthis lecture!! SQL will frequently be used duringthis lecture!! SQL will frequently be used duringthe semester, even though we are going to learnthe semester, even though we are going to learna different relational language!a different relational language!

Material for selfMaterial for self--studystudy(in case your background is(in case your background isweak, dated, or missing):weak, dated, or missing):

•• Extra Extra slidesslidesvia IISvia IIShomepagehomepage

•• Cheap and easy Cheap and easy tutorialstutorialsfrom the Schaumfrom the Schaum‘‘ s seriess series


MotivationMotivation

MotivationMotivationMotivation


WS 2016/17WS 2016/17


Intelligent Intelligent DatabaseDatabase or Intelligent or Intelligent Database SystemDatabase System??

DBDBDBMSDBMS

. . . .. . . .

DBSDBS

??

Where is „intelligence“ located?In the DBor in the DBMS?

Or even outside the DBS?

Where is Where is „„ intelligenceintelligence““ located?located?In the In the DBDB or in the or in the DBMSDBMS??

Or even outside the DBS?Or even outside the DBS?


IDBS: IDBS: „„ Intelligent ServicesIntelligent Services““ in a DBMSin a DBMS

DBDBDBMSDBMS

DBSDBS

genericgenericspecificspecific

Certainly required: Certainly required: „„ IntelligentIntelligent““ behaviour of the DBS, behaviour of the DBS, i.e.,i.e., generic (applicationgeneric (application--independent) servicesindependent) servicesinside the DBMSinside the DBMS, able to , able to „„ simulate intelligencesimulate intelligence““


IDBS: IDBS: „„ KnowledgeKnowledge““ Inside a DBInside a DB

DBDBDBMSDBMS

DBSDBS

genericgeneric

specificspecific

Also certainly required: Also certainly required: „„ KnowledgeKnowledge““ about the resp. application domain about the resp. application domain in the DDin the DD(Data Dictionary)(Data Dictionary)

„„ KnowledgeKnowledge““ : : RulesRulesfrom the application domain as a basis forfrom the application domain as a basis fordrawing intelligent conclusions from stored datadrawing intelligent conclusions from stored data

DDDD


IDBS: IDBS: „„ TraditionalTraditional““ Approach with External System ComponentsApproach with External System Components

DBDBDBMSDBMS

DBSDBS

Inference System, Agent System, Expert SystemInference System, Agent System, Expert System

Knowledge Base,Knowledge Base,Rule BaseRule Base

„„ loose couplingloose coupling““

Preferred by many: Move Preferred by many: Move „„ IntelligenceIntelligence““ and and „„ KnowledgeKnowledge““ out ofout of the DBSthe DBS


IDBS: Our Approach IDBS: Our Approach –– „„ In databaseIn database--IntelligenceIntelligence““ !!

DBDBDBMSDBMS

DBSDBS

Approach favoured by our research group (and thus in this lecturApproach favoured by our research group (and thus in this lecture):e):•• Try to reach as much Try to reach as much „„ intelligenceintelligence““ as possible using as possible using existing DB technologyexisting DB technology!!•• Identify weaknesses of this technology and think about reasoIdentify weaknesses of this technology and think about reasonablenable

extensionsextensions, without leaving the DB context! , without leaving the DB context!

„„ tight couplingtight coupling““


At the core of IIS: Theory and Practice of Deductive DatabasesAt the core of IIS: Theory and Practice of Deductive Databases

This approach This approach –– which is a special one which is a special one ––explains the drawing on the title slide ofexplains the drawing on the title slide ofthis lecture.this lecture.

Therefore:Therefore:Theory and Practice of the establishedTheory and Practice of the establishedresearch area of research area of „„ Deductive DatabasesDeductive Databases““will be at the core of this lecture.will be at the core of this lecture.

The essence of this area of research can beThe essence of this area of research can bedescribed as follows:described as follows:

How to analyse data using stored queriesHow to analyse data using stored queries(in SQL: (in SQL: viewsviews) that serve as declarative) that serve as declarativeanalytical programs?analytical programs?


Datalog and SQLDatalog and SQL

SQL:SQL: Datalog:Datalog:

•• used in used in industryindustryand commerceand commerce•• supported by many DBMS supported by many DBMS productsproducts•• standardizedstandardized•• useruser--friendlyfriendly ((„„ controlled Englishcontrolled English““ ))•• rich rich set of syntactic featuresset of syntactic features

•• used in used in academiaacademiaonlyonly•• just few academic just few academic protoypesprotoypes•• nono standardsstandards•• mathematicalmathematicalstylestyle•• minimalisticminimalisticsyntaxsyntax

•• Research in deductive databases has a nearly 40Research in deductive databases has a nearly 40--years history (as old as SQL), butyears history (as old as SQL), buthas been using a has been using a differentdifferentdeclarative language (not SQL!) most of the time, stronglydeclarative language (not SQL!) most of the time, stronglyinfluenced by the logic programming language PROLOG:influenced by the logic programming language PROLOG:

•• Nearly all publications in this area have been using Nearly all publications in this area have been using DatalogDatalog–– thatthat‘‘ s why we will uses why we will useDatalog during this lecture, too (and you will have to learnDatalog during this lecture, too (and you will have to learnit!).it!).

•• Many results of DDB research have been transferred to the Many results of DDB research have been transferred to the SQLSQL world recently!world recently!ThatThat‘‘ s why SQL will also be appearing throughout the lecture in varios why SQL will also be appearing throughout the lecture in various places.us places.

DatalogDatalog


Datalog vs. SQL: Comparison in a NutshellDatalog vs. SQL: Comparison in a Nutshell

s(X) ← p(X,Y).s(X) ← r(Y,X).

t(X,Y,Z) ← p(X,Y), r(Y,Z).

w(X) ← s(X), not q(X).

s(X) s(X) ←← p(X,Y).p(X,Y).s(X) s(X) ←← r(Y,X).r(Y,X).

t(X,Y,Z) t(X,Y,Z) ←← p(X,p(X,YY), r(), r(YY ,Z).,Z).

w(X) w(X) ←← s(X), s(X), notnot q(X).q(X).

Datalog rulesDatalog rulesDatalog rules CREATE VIEW s ASCREATE VIEW s AS(SELECT a FROM p)(SELECT a FROM p)

UNIONUNION(SELECT b FROM r);(SELECT b FROM r);

CREATE VIEW t ASCREATE VIEW t ASSELECT a, b, cSELECT a, b, cFROM p, rFROM p, rWHERE WHERE p.b = r.ap.b = r.a,,

CREATE VIEW w ASCREATE VIEW w AS(TABLE s)(TABLE s)

MINUSMINUS(TABLE q); (TABLE q);

SQL viewsSQL viewsSQL views


Datalog Basics on a Single SlideDatalog Basics on a Single Slide

p(1,a).p(2,b).p(3,c).

q(2).q(5).

r(a,1).r(a,2).r(b,3).

p(1,a).p(2,b).p(3,c).

q(2).q(5).

r(a,1).r(a,2).r(b,3).

s(X) ← p(X,Y).s(X) ← r(Y,X).

t(X,Y,Z) ← p(X,Y), r(Y,Z).


s(X) ← p(X,Y).s(X) ← r(Y,X).

t(X,Y,Z) ← p(X,Y), r(Y,Z).


FactsFactsFacts

RulesRulesRules

ConstantsConstantsConstants

ConjunctionConjunctionConjunction

NegationNegationNegation

p, q, r: Base relations s, t, w: Derived relations

Variables Variables Variables

Relation NamesRelation NamesRelation Names


Structure of the CourseStructure of the Course

1.1. Organisation and MotivationOrganisation and Motivation 1 lecture1 lecture

2.2. Deduction in Datalog and SQL Deduction in Datalog and SQL 7 lectures7 lectures3.3. Semantics of Deductive DatabasesSemantics of Deductive Databases 6 lectures6 lectures4.4. Efficient Query Evaluation in DDBsEfficient Query Evaluation in DDBs 6 lectures6 lectures5.5. Efficient Update Propagation in DDBsEfficient Update Propagation in DDBs 7 lectures7 lectures

6.6. PerspectivesPerspectives 1 lecture1 lecture

relevant forrelevant forexamexam

This is how the lecture will be structured This is how the lecture will be structured –– the number of lectures might be slightly varying the number of lectures might be slightly varying in in „„ real lifereal life““

Timetable:Timetable: •• Chapter 1: todayChapter 1: today•• Chapter 2: Oct/NovChapter 2: Oct/Nov•• Chapter 3: Nov/DecChapter 3: Nov/Dec•• Chapter 4: Dec/JanChapter 4: Dec/Jan•• Chapter 5: Jan/FebrChapter 5: Jan/Febr•• Chapter 6: 8.2.2017Chapter 6: 8.2.2017

Chapter 5 is new this year!Chapter 5 is new this year!Chapter 5 is new this year!

Date post:	26-May-2018
Category:	Documents
Upload:	vuongtu
View:	220 times
Download:	1 times

Intelligent Information Systems - uni-bonn.de© 2016 Prof. Dr. Rainer Manthey Intelligent...

Documents