AMTI – Knowledge Management
[email protected]
Knowlegde • Knowledge = contextual, relevant, actionable information – – – – – – –
Strong experiential and reflective elements Good leverage and increasing returns Dynamic Branches and fragments with growth Difficult to estimate impact of investment Uncertain value in sharing Evolves over time with experience
One Perspective of KM • “KM [Knowledge Management] involves blending a company’s internal and external information and turning it into actionable knowledge via a technology platform.” Susan DiMattia and Norman Oder in Library Journal, September 15, 1997.
Two Kinds of Knowledge • Tacit: or unarticulated knowledge is more personal, experiential, context specific, and hard to formalize; is difficult to communicate or share with others; and is generally in the heads of individuals and teams. • Explicit: explicit knowledge can easily be written down and codified.
KM: Learning and Communication Process • In simple language KM is an effort to capture not only explicit factual information but also the tacit information and knowledge that exists in an organization, usually based on the experience and learning of individual employees, in order to advance the organization's mission. • The eventual goal is to share knowledge among members of the organization.
Where does KM come from? • Technology – Infrastructure, Database, Web, Interface
• Globalization – World wide markets, North American integration
• Demographics – Aging population, workforce mobility, diversity
• Economics – Knowledge about economy
• Customer relations – Quality
• Increase in information – Specialization, Volume, Order
KM: Organizational Learning • Ability to learn from past • To improve, organization must learn • Issues – Meaning, management, measurement
• Activities – Problem-solving, experimentation, learning from past, learning from acknowledged best practices, transfer of knowledge within organization
• Must have organizational memory, way to save and share it • Organizational learning – Develop new knowledge
• Organizational culture – Pattern of shared basic assumptions
Knowledge Management • Aims – Make knowledge visible – Develop knowledge intensive culture – Build knowledge infrastructure
• Processes – Creation of knowledge – Sharing of knowledge – Seeking out knowledge – Using knowledge
KM Processes • Knowledge creation – Generating new ideas, routines, insights – Modes • Socialization, externalization, internalization, combination
• Knowledge sharing – Willing explanation to another directly or through an intermediary
• Knowledge seeking – Knowledge sourcing
Knowledge
Knowledge and IT
Knowledge Base • The first step in constructing an AI program is to build a knowledge base • Will be used by the inference mechanism to reason and draw conclusions Computer
Inputs
Knowledge base
Inference mechanism
Outputs
Knowledge Base • Knowledge engineering: process of collecting and organizing the knowledge • Knowledge representation: process of how knowledge is represented to form a knowledge base
Representasi Pengetahuan • Bagaimana merepresentasikan pengetahuan ke dalam basis pengetahuan dan menguji kebenaran penalaran • Cara-cara lama: – List, digunakan pada LISP – Predicate Calculus, digunakan pada Prolog – Tree, untuk heuristic search
• Karakteristik RP: – Dapat diprogramkan – Dapat dimanfaatkan untuk penalaran, menggambarkan kesimpulan sebagai fungsi kecerdasan
Alasan Pemilihan • Why knowledge representation rather than information representation? – Karena pada konvensional database merepresentasikan data secara sederhana: string, number, boolean – Namun AI menganggap pengetahuan lebih kompleks, seperti proses, prosedur, aksi, waktu, tujuan dan penalaran
Representasi Pengetahuan (2) • Harus terdiri dari struktur data dan prosedur untuk penafsiran • Hal yang berhubungan dengan RP: – Object pengetahuan itu sendiri – Event: kejadian-kejadian dalam dunia nyata dan hubungannya – Performa: bagaimana melakukan suatu tugas tertentu – Meta knowledge: pengetahuan tentang pengetahuan yang direpresentasikan
Penggunaan Pengetahuan • Acuisition: mengintegrasikan informasi baru kedalam pengetahuan sistem. – Dua level: • Menyusun fakta ke dalam database • Pembuatan fungsi untuk mengintegrasikannya dengan cara “belajar dan mengadaptasikannya” terlebih dahulu
• Retrieval: mengingat kembali, menyusun ulang pengetahuan berdasarkan hubungan pengetahuan terhadap masalah – Linking: mengekstrak informasi baru tersebut – Lumping: mengelompokkan hasil ekstraksi pengetahuan baru tersebut kedalam struktur yang lebih besar seperti yang dibutuhkan dalam menyelesaikan masalah
Penggunaan Pengetahuan (2) • Reasoning: pengetahuan digunakan untuk menalar suatu permasalahan – Formal reasoning: menggunakan logika proporsional – Procedural reasoning: menggunakan aturan produksi ( IF-THEN) – Analogical reasoning: sangat sulit
Klasifikasi Kategori RP • Menurut Mylopoulus dan Levesque: (declarative) – Representasi Logika: menggunakan logika formal. Digunakan pada PROLOG – Representasi Prosedural: menggambarkan prosedur sebagai kumpulan instruksi untuk memecahkan masalah. Digunakan dalam pemrograman: IF-THEN – Representasi Network: menggambarkan pengetahuan sebagai Graph dan Tree – Representasi Terstruktur: memperluas konsep Representsi Network dengan membuat nodenodenya menjadi struktur data yang kompleks. Contoh: script, frame, dan object
List dan Tree • List:serangkaian struktur data yang dibuat secara berhubungan, list bisa juga menggambarkan relasi dan hirarki • Tree: suatu struktur data yang berupa node-node yang dibuat secara hirarkis dan hubungannya • Lihat di Struktur Data!
Lists & Trees (2) List:
Tree: Node
Arc
Sematic Network • Diperkenalkan oleh Ros Quillian • Very flexible: almost any kind of object, attribute, concept, etc. can be defined and relationship created with links • To seek answer: the computer simply searches forward or backward through the arcs from a starting node • Terdiri dari: – lingkaran-lingkaran yang menunjukkan obyek dan informasi mengenai obyek tersebut – panah (arc) yang menunjukkan hubungan antar obyek • Kelebihan: – Memiliki sifat inheritance • Menggunakan representasi OAV (Object Atributte Value)
Semantic Network (2)
Frame • Diperkenalkan oleh Minsky tahun 1975 • Suatu struktur data yang digunakan untuk merepresentasikan pengetahuan dan situasi-situasi yang telah dipahami • Frame memiliki slot untuk menggambarkan rincian dan karakteristik obyek
Frames (2)
Script • Mirip dengan frame, merepresentasikan pengetahuan berdasarkan pengalamanpengalaman • Frame menggambarkan obyek, sedangkan script menggambarkan urutan peristiwa • Elemen script: – – – – – –
Kondisi input: start, awal Track: variasi yang mungkin terjadi Prop: obyek pendukung Role: peran yang dimainkan oleh suatu obyek Scène: adegan yang terjadi Hasil (result): kondisi akhir yang terjadi
Schemas: Scripts (2)
Knowledge-based systems
Structure and characteristics 1 • KBSs are computer systems – contain stored knowledge – solve problems like humans would
• KBSs are AI programs with program structure of new type – knowledge-base (rules, facts, meta-knowledge) – inference engine (reasoning and search strategy for solution, other services)
• characteristics of KBSs: – intelligent information processing systems – representation of domain of interest → symbolic representation – problem solving → by symbol-manipulation
Structure and characteristics 2
Main components 1 • knowledge-base (KB) – knowledge about the field of interest – symbolically described system-specification – KNOWLEDGE-REPRESENTATION METHOD!
• inference engine – „engine” of problem solving (general problem solving knowledge) – supporting the operation of the other components – PROBLEM SOLVING METHOD!
• case-specific database – auxiliary component – specific information (information from outside, initial data of the concrete problem) – information obtained during reasoning
Main components 2 • explanation subsystem explanation of system’ actions in case of user’ request typical explanation facilities: – explanation during problem solving: • WHY... (explanative reasoning, intelligent help, tracing information about the actual reasoning steps) • WHAT IF... (hypothetical reasoning, conditional assignment and its consequences, can be withdrawn) • WHAT IS ... (gleaning in knowledge-base and case-specific database)
– explanation after problem solving: • HOW ... (explanative reasoning, information about the way the result has been found) • WHY NOT ... (explanative reasoning, finding counter-examples) • WHAT IS ... (gleaning in knowledge-base and case-specific database)
Main components 3 • knowledge acquisition subsystem – main tasks: • • • • •
checking the syntax of knowledge elements checking the consistency of KB (verification, validation) knowledge extraction, building KB automatic logging and book-keeping of the changes of KB tracing facilities (handling breakpoints, automatic monitoring and reporting the values of knowledge elements)
• user interface (→ user) – dialogue on natural language (consultation/ suggestion)
• specially intefaces – database and other connections
• developer interface (→ knowledge engineer, human expert)
Main components 4 • the main tasks of the knowledge engineer: – knowledge acquisition and design of KBS: determination, classification, refinement and formalization of methods, thumb-rules and procedures – selection of knowledge representation method and reasoning strategy – implementation of knowledge-based system – verification and validation of KB – KB maintenance
Perbedaan Pengguna KBS • Manager: apa yang dapat saya gunakan? • Teknolog: bagaimana saya dapat mengimplentasikan teknologi dengan baik? • Peneliti: bagaimana saya dapat mengembangkannya • User: bagaimana dapat membantu saya? Dapat menghemat biaya? Bagaimana kehandalannya?
Example of KBS • The famous: – MYCIN: diagnosa penyakit, – DENDRAL: mengidentifikasi struktur molekul campuran kimia yang tidak dikenal, – XCON & XSEL: konfigurasi sistem komputer besar, – Prospector: bidang geologi
• The other: – SOPHIE: analisis sirkuit elektronik, – DELTA: pemeliharaan lokomotif listrik, – FOLIO: stok dan investasi
Benefits of KBS • Memungkinkan orang awam dapat mengerjakan pekerjaan para ahli • Bisa melakukan proses berulang secara otomatis • Menyimpan pengetahuan dan keahlian pakar • Meningkatkan output dan produktifitas • Melestarikan keahlian pakar • Dapat beroperasi pada lingkungan berbahaya • Dapat meningkatkan kemampuan sistem komputer • Dapat bekerja dengan informasi yang tidak lengkap • Sebagai media pelengkap dalam pelatihan • Menghemat waktu pengambilan keputusan
The Down Side of KBS • Development of an KBS is difficult • KBS is expensive • Most KBS still must be implemented & delivered on a big mainframe or minicomputer • Not 100% reliable • Kepakaran tidak selalu tersedia pada bidang-bidang tertentu
KBS Categories • KBS is not suitable for all situations • Generic KBS categories: – – – – – – – –
Control : intelligent automation Debugging : recommends corrections to faults Design : developing products to specification Instruction : optimized computer instruction Interpretation : clarification of situations Planning : developing goal-oriented schemes Prediction : intelligent guessing of outcomes Repair : automatic diagnosis, debugging, planning and fixing
NEXT • Presentasi – Senin, 5 Desember 2011 – Pukul 08.00
