Transcript 배운봉-study_-_collective_intelligence，_Social_network

Collective Intelligence
By Yunfeng Pei
2010-04-17
•
•
Prof. I. J. Chung
Dept. of Computer & Information Science, Korea Univ.
1
Collective Intelligence
Concept
집단지성(集体智慧/collective intelligence)란 다수의 개체들
이 서로 협력 혹은 경쟁을 통하여 얻게 되는 지적 능력에 의
한 결과로 얻어진 집단적 능력을 말한다.
집단적 지적 능력을 통해 개체적으로는 미미하게 보이는 동
물, 사람의 능력을 모으는 과정을 통한 결정 능력의 다양한
형태로 한 개체의 능력 범위를 넘어선 힘을 발휘할 수도 있
다고, 이 분야는 사회학, 경영학, 컴퓨터 공학 등에서 연구되
고 있다.
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
2
Collective Intelligence
History
1910년대 하버드 대학 교수이자 곤충학자인 William Morton
Wheeler 가 개미의 사회적 행동을 관찰하면서 처음 제시했
다.
휠러는 개체로는 미미한 개미가 공동체로서 협업(協業)하여 거대한 개미집을 만
들어내는 것을 관찰하였고, 이를 근거로 개미는 개체로서는 미미하지만 군집(群
集)하여서는 높은 지능체계를 형성한다고 설명하였다.
Pierre Levy는 사이버 공간의 집단지성을 제시하였다.
그는 "누구나 자신의 공간 ( 사이트 ) 를 가지고 일종의 형성하는 시대가 오면 어
디에나 분포하고, 지속적으로 가치 부여되며, 실시간으로 조정되고, 역량의 실제
적 동원에 이르는 집단지성이 발현될 것"이라고 주장하였다.
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
3
Collective Intelligence
Four principles
CI 는 대규모 협조이다. 이런 계념을 실현하기 위해서는 4가
지 기본 원칙이 필요하다.
Openness
Peering (对等)
Sharing
Acting globally
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
4
Collective Intelligence
Types
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
5
Collective Intelligence
Examples
Wikipedia
Wikipedia 의 발전 과정은 지식·정보의 생산자나 수혜자가 따로 없이 누구나 생산
할 수 있고 모두가 손쉽게 공유하면서도 정체되지 않고 계속 진보하는, 집단지성
의 특성을 보여준다.
Google search
웹 페이지 평가에 다른 페이지에서 얼마나 링크를 많이 받았는지를 적용한 첫 검
색엔진이다. 수천 명이 특정 웹 페이지에 대해 말한 정보를 랭킹 방법에 적용하여
검색 결과 순서를 정하는 데 이용했다.
Web 2.0
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
6
Collective Intelligence
Benefits of collective intelligence
Higher retention rates
The more users interact with the application, the stickier it
gets for them, and the higher the probability that they’ll
become repeat visitors.
Greater opportunities to market to the user
The greater the number of interactions, the greater the
number of pages visited by the user, which increases the
opportunities to market to or communicate with the user.
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
7
Collective Intelligence
Benefits of collective intelligence
Higher probability of a user completing a transaction
and finding information of interest
The more contextually relevant information that a user finds, the
better the chances that he’ll have the information he needs to
complete the transaction or find content of interest. This leads to
higher click-through and conversion rates for your advertisements.
Boosting search engine rankings
The more users participate and contribute content, the
more content is available in your application and indexed
by search engines. This could boost your search engine
ranking and make it easier for others to find your
application.
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
8
Collective Intelligence
CI in web applications
Three things that need to happen to apply collective intelligence in your
application.
Allow users to interact with your site and with each other, learning
about each user through their interactions and contributions.
Aggregate what you learn about your users and their contributions
using some useful models.
Leverage those models to recommend relevant content to a user.
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
9
Collective Intelligence
CI in web applications
Three components to harnessing collective intelligence.
Allow users to interact.
Learn about your users in aggregate.
Personalize content using user interaction data and
aggregate data.
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
10
Social Network
By Yunfeng Pei
2010-04-17
•
•
Prof. I. J. Chung
Dept. of Computer & Information Science, Korea Univ.
11
Social Network
Concept
Social Network 는 web science 의 연구 분야 중 하나로,
web 상에서 개인 또는 집단이 하나의 노드(node)가 되어 각
노드들 간의 상호의 존적인 관계(tie)에 의해 만들어지는 사
회적 관계 구조를 말한다.
Node 는 network에서 존재하는 개별적인 주체들이다.
Tie 는 각 node 들의 관계를 말한다.
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
12
Social Network
Social network analysis - SNA
Social network analysis 사람과 사람들 간의 중요한 지식관
계를 반영하며, 특별히 조직간의 협조를 향상시키고, 지식을
창조하며,지식을 전파한다.
Social network analysis 는 수많은 노드들과 그 노드들 사이
의 무수히 다양한 관계들로 인해 계산론적으로 접근하기에
매우 복잡한 분야이다.
Social network analysis 는 현재 중요한 기술로 거듭나고 있
다. 그리고 또한 이슈되는 연구분야이다.
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
13
Social Network
Social network analysis
Solve some problems
Choose a leader
어떠한 사람이 수많은 사람들 중에서 신뢰성이 높고 존경을 받는 것인가 ?
Choose a task team
어떻게 하면 조직네에서 연결이 있는 사람들을 하나의 팀으로 조성하는가 ?
Mergers & Acquisitions
두 문화가 합병되는 것만이 아니라, 두 독립적인 네트워크를 합병하는 것이다.
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
14
Social Network
An example of a social network diagram.
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
15
Social Network
Social network analysis
Betweenness
A node lies between other nodes in the network.
The connectivity of the node's neighbors.
Giving a higher value for nodes which bridge clusters.
Bridge
An edge is said to be a bridge if deleting it would cause its
endpoints to lie in different components of a graph.
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
16
Social Network
Social network analysis
Closeness
The degree an individual is near all other individuals in
a network.
It reflects the ability to access information.
Thus, closeness is the inverse of the sum of the
shortest distances between each individual and every
other person in the network.
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
17
Social Network
Social network analysis
Degree
The count of the number of ties to other actors in the
network.
Network or global-level density is the proportion of
ties in a network relative to the total number possible.
The degree that a node contributes to sum of
maximum flow between all pairs of nodes.
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
18
Social Network
Social network analysis
Centralization
The difference between the number of links for each
node divided by maximum possible sum of differences.
A centralized network will have many of its links
dispersed around one or a few nodes, while a
decentralized network is one in which there is little
variation between the number of links each node
possesses.
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
19
Social Network
Social network analysis
Social network analysis software
Network analytic tools are used to represent the
nodes (agents) and edges (relationships) in a
network, and to analyze the network data.
Network analysis tools allow researchers to
investigate large networks like the Internet, disease
transmission, etc.
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
20
The Semantic Web
By Yunfeng Pei
2010-04-17
•
•
Prof. I. J. Chung
Dept. of Computer & Information Science, Korea Univ.
21
The Semantic Web
Concept
The Semantic Web is an evolving development of the
World Wide Web in which the meaning of information
and services on the web is defined, making it possible
for the web to "understand" and satisfy the requests of
people and machines to use the web content.
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
22
The Semantic Web
The difference
The world wide web
Hypertext Markup Language (HTML).
HTML 로 작성된 contents 는 컴퓨터가 해석할수 있는 meta data 보다는 사람의
시각정보에 대한 meta data와 자연어로 작성된 contents 로 되여있다.
The semantic web
Resource Description Framework (RDF),
Web Ontology Language (OWL),
Extensible Markup Language (XML).
이런 언어를 기반으로 작성된 contents는 컴퓨터가 해석할수 있는 meta data로
구성.
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
23
The Semantic Web
Factor
Ontology
Ontology 는 일종의 지식표현으로서 컴퓨터를 하여끔
ontology 로 표현된 개념을 이해하고 지식처리를 할 수 있는것
이다.
agent
유저를 대신햐여 정보를 수집하고, 정보를 검색하며, ontology를 이용
하여 다른 agent과 교류하여 유저대신에 일상 일를 처리하는 것이다.
agent는 the semantic web 중에서 중요한 요소이다.
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
24
The Semantic Web
Components
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
25
The Semantic Web
Components
XML
Extensible Markup Language. The language framework that
has been used to define nearly all new languages that are
used to interchange data over the web.
XML Schema
A language used to define the structure of specific XML
languages.
RDF
A flexible language capable of describing all sorts of
information and meta data.
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
26
The Semantic Web
Components
RDF Schema
A framework that provides a means to specify basic
vocabularies for specific RDF application languages to use.
RDF Schema is a framework for constructing ontologies and
is used by many more advanced ontology frameworks.
Ontology
Languages used to define vocabularies and establish the
usage of words and terms in the context of a specific
vocabulary.
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
27
The Semantic Web
Components
Logic and Proof
Logical reasoning is used to establish the consistency and
correctness of data sets and to infer conclusions that aren’t
explicitly stated but are required by or consistent with a
known set of data. Proofs trace or explain the steps of
logical reasoning.
Trust
A means of providing authentication of identity and evidence
of the trustworthiness of data, services, and agents.
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
28
The Semantic Web
Vision
The semantic web is a vision of information that is
understandable by computers, so that users can perform
more of the tedious work involved in finding, combining,
and acting upon information on the web.
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
29
The Semantic Web
Application
RSS(RDF Site Summary)
Web blog
Portal
Search Engine
The semantic web service
Business Intelligence
Knowledge Management
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
30
Ontology
By Yunfeng Pei
2010-04-17
•
•
Prof. I. J. Chung
Dept. of Computer & Information Science, Korea Univ.
31
Ontology
Concept
In theory
Ontologies are defined as a formal specification of a
shared conceptualization.
In English
An ontology provides a shared vocabulary, which can be
used to model a domain, that is, the type of objects and/or
concepts that exist, and their properties and relations.
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
32
Ontology
Components
Class
sets, collections, concepts, classes in programming, types of
objects, or kinds of things.
Instances
instances or objects
Relation
ways in which classes and individuals can be related to one
another
Property
aspects, properties, features, characteristics, or parameters
that objects (and classes) can have
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
33
Ontology
Ontology language
RDF (Resource Description Framework)
RDF is a standard model for data interchange on the Web.
RDF has features that facilitate data merging even if the
underlying schemas differ, and it specifically supports the
evolution of schemas over time without requiring all the
data consumers to be changed.
OWL (Web Ontology Language)
The W3C Web Ontology Language (OWL) is a Semantic
Web language designed to represent rich and complex
knowledge about things, groups of things, and relations
between things.
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
34
Ontology
Ontology language
SWRL (Semantic Web Rule Language)
SWRL is a proposal for a Semantic Web rules-language,
combining sublanguages of the OWL Web Ontology
Language (OWL DL and Lite) with those of the Rule Markup
Language.
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
35
Ontology
Perspectives
Library and Information Science
Artificial Intelligence
Natural Language Processing
The Semantic Web
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
36
Ontology
Artificial Intelligence
Objectives
Model common sense and domain knowlege
Usage
Knowledge representation and reasoning
Examples
CYC,…
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
37
Ontology
The Semantic Web
Objectives
Provide semantics for web resource
Usage
Describe resources and their contents
Examples
DC, DAML-library,…
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
38
Ontology
The Ontology Web Language – OWL
Goal
Formally describe the semantics of classes and
properties used in web documents.
Go beyond the basic semantics in RDFS
Current status
Use cases and their requirements on ontologies
Eight design goals
Requirements
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
39
Ontology
The Ontology Web Language – OWL
Design goals
Shared ontologies
Ontology interoperability
Inconsistency detection
Balance of expressivity and scalability
Ease of use
XML syntax
Internationalisation
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
40
Ontology
The Ontology Web Language – OWL
Requirements
Ontologies as distinct objects
Unambiguous term referencing with URIs
Explicit ontology extension
Commitment to ontologies
Ontology metadata
Versioning information
Class definition primitives
Property definition primitives
Datatypes
Class and property equivalence
Local unique names assumption
Etc….
2015-04-13
• Prof. I. J. Chung
•Dept. of Computer & Information Science, Korea Univ.
41