[Summary] Ontologies for Crisis Management : A review of State of the Art in Ontology Design and Usability – ISCRAM 2013

by Shuangyan Liu, Christopher Brewster, Ducan Shaw

This paper presents a review of ontologies, vocabularies and taxonomies useful in crisis management systems.

– Analyse different ontologies and vocabularies available

– Propose 11 subject areas

Drawback of current works :

Current literature fails to provide an overall picture of the subject areas involved in crisis management and how the concepts should be represented in crisis information systems.

11 subject areas

  1. Resources
  2. Processes
  3. People
  4. Organisations
  5. Damage
  6. Disasters
  7. Infrastructure
  8. Geography
  9. Hydrology
  10. Meteorology
  11. Topography

————————————

Semantic interoperability refers to the ability of computer systems to communicate data with a unified meaning

From the perspective of information management, crisis management is a complex domain with many pieces of information and corresponding concepts involved before, during and after a crisis

Advertisements

[Summary] A Metamodel and its Ontology to Guide Crisis Characterization and its Collaborative Management – ISCRAM 2008

by Frédérick Bénaben, Chihab Hanachi, Matthieu Lauras, Pierre Couget, Vincent Chapurlat (2008)

MIS Design proposal in the MDA context :

1. Business Branch : From Knowledge about Crisis through Inference Mechanism => CIM (BPMN Models) through Model Transformation => PIM (UML or DSL Model)

2. Technology Branch : From Knowledge about IS Architecture through Technological Platform Modeling => PM (Agile ESB Model)

3. Business Branch + Technology Branch through Technological Modeling => PSM (MIS Model)

 [ISyCri Project] Risk is the possibility of the occurence of an event having positive or negative consequences

  1. Its probability
  2. The measure of its potential effects

5 types of risks (according to PRIM – a French institutional website for major risks prevention):

  1. Natural risks (landslide, fire, flood, hurricane, earthquake, etc.)
  2. Technological risks (industrial accident, nuclear accident, biological accident, etc)
  3. Collective transport risks (people or material)
  4. Everyday life risks (car accident, accident at home, etc)
  5. Risks linked to conflicts

 

Glossary

MIS : Mediation Information System

IS : Information Systems

SoS : System of Systems

CIM : Computational Independent Model  (BPMN model of collaborative process)

PIM : Platform Independent Model (UML or DSL model of Mediation Information System)

PM : Platform Model (Agile ESB model)

PSM : Platform Specific Model = MIS model ready to use

MDA : Model Driven Architecture (a model-based approach to design system, which separate business and technolgoy)

[Summary] Using a Workflow Management System to Manage Emergency Plans

by Christian Sell and Iris Braun (2009)

I. Problem to be addressed

Propose to use workflows in Emergency Response for supporting modeling, execution and management

II. Insufficiency of existing works

Emergency plans describe the chronological order, only exist in the form of large printed documents => Technical support for executing is very limited

III. Contributions

  1. It shows how to use the workflow to manage the emergency plan
  2. It propose a meta-model of workflow to execute the emergency plan

Một văn bản viết ra, quy định nhiệm vụ cho các đơn vị liệu có chồng chéo nhau ? Làm cách nào để kiểm tra những chồng chéo này ? Kiểm tra văn bản viết ra có đúng và đầy đủ ?

Từ văn bản viết ra, sẽ có các cuộc diễn tập nhằm hoàn thiện văn bản, tại sao phải tốn tiền, vật lực vào các cuộc diễn tập quy mô lớn ?

Workflow cho phép mô phỏng các nhiệm vụ, tác nhân trên máy tính, giảm thiểu tiền bạc

[Summary] Coordination of Emergency Response Operations via the Event-Based Awareness Mechanism

by Bo Yu and Guoray Cai, College of Information Sciences and Technology, The Pennsylvania State University (2012)

The idea is to provide actors (agent) an awareness of their current situation and each other’s activities so that they can make mutual adjustment in coordination processes = > COORDINATION by AWARENESS

The author believes that an Event-Driven Model of coordination is most appropriate to the domain of emergency response

The model presented in this article does not rely on a pre-determined process model. Instead, it is based on a Situational Representation of activities and dependencies and uses this situated knowledge to perform event reasoning

Based on the definition of coordination of [Malone and Crowston 1994], dependency becomes the central concept

Activity-centered approach allows actors to create new activities and dependencies during the process

Some types of dependencies between activities :

  1. Resource-related dependencies : consider the management of shared resources involved in multiple activities
  2. Goal-related dependencies :
  3. Constraint-related dependencies : reflect the temporal and spatial constraints between multiple activities

[Summary] An Ontology for Coordination

by Ben Lithgow Smith, Valentina Tamma, and Michael Wooldridge (Department of Computer Science, The University of Liverpool, United Kingdom)

This article describes an ontological approach to coordination in which agents dynamically manage the interdependencies that arise during their interactions.

Aim of this article : Provide coordination at run time rather than being hard-wired at design time.

Coordination : How to manage interdependencies between the activities of agents ?

[Summary] 2012 Presidential Elections on Twitter – An Analysis of How the US and French Election were Reflected in Tweets

by Farhad Nooralahzadeh, Viswanathan Arunachalam, Costin Chiru (Politehnica University of Bucharest, Department of Computer Science and Engineering)

TWITTER DATASET

Tweets concerning US 2012 election were crawled in October and November, 2012. The dataset was obtained following the 3 steps :

  1. We obtained all the tweets that contained the candidate names, candidate addresses, election related hash tags and candidates hash tags (by crawling the Twitter using Twitter API). We collected the full text, the author, the written date and time, and we ended up with 196,000 tweets and stored them in a database.
  2. We cleaned each tweet by removing stop words, numbers, html references, punctuation symbols and candidate names and addresses. Furthermore we deleted all the re-tweets from the dataset (inorder to remove duplicate tweets). Emotions were found and replaced by the word posemo for positive emotions and negemo for negative emotions. We lemmatized words in the tweets by using TreeTagger and stored the cleaned and lemmatized tweet beside the original one in the database.
  3. We also extracted the candidate names from each tweet and stored it in other column. If more than one candidate was present in the text of the tweet, we marked it with the value +1 candidate. We also marked the tweets where no candidate was mentioned with the value no candidate.

[Summary] Coordination Techniques for Distributed Artificial Intelligence (1996)

by N.R. Jennings

Nêu vấn đề phối hợp (tại sao phải phối hợp ???),

Mô hình hóa hệ thống AI phân tán như một vấn đề tìm kiếm mục tiêu phân tán

(Modelling Distributed AI Systems as a Distributed Goal Search Problem)

Mô hình Commitment và Convention của phối hợp

Những kỹ thuật phối hợp chung và cách sử dụng Commitment, Conventions

This model of coordination is founded upon the hypothesis : All coordination mechanisms can ultimately be reduced to commitments and their associated (social) conventions

Commitments and Conventions (means of monitoring commitments in changing circumstance) are the foundation of coordination in all DAI systems.

Inter-dependencies can be classified along 2 orthogonal dimensions :

  1. whether they are weak or strong
  2. whether they are uni-directional or bi-directional

Joint actions (in which a team of agents decide to pursue a common goal in cooperative manner) can be characterized as having the following properties :

  1. the team members are mutually responsive to one another
  2. the team members have a joint commitment to the joint activity
  3. the team members are committed to be mutually supportive of one another during the pursuit of their joint objective

Joint goals differ from individual goals in that they are not directly associated with actions – for this reason, they must be mapped onto individual goals as only individual agents have the ability to act (perform domain level tasks)

There are 3 main reasons why the actions of multiple agents need to be coordinated :

  1. there are dependences between agents’ actions
  2. there is a need to meet global constraints
  3. no one individual has sufficient competence, resources or information to solve the entire problem

3 prominent coordination techniques :

  1. Organisational Structuring : can be viewed as a pattern of information and control relationships between individuals. These control relationships, be they hierarchical, heterarchical or flat, are responsible for designating the relative authority of the agents and for shaping the types of social interaction which can occur.
  2. Meta-level information exchange : involves agents sending each other control level information about their current priorities and focus. Meta-level information exchange is a medium term source of knowledge about an agent’s commitments – shorter than organisation structures but longer than multi-agent planning approaches
  3. Mutl-agent planning : agents usually form a plan which specifies all their future actions and interactions with respect to achieving a particular objective. It details, before execution commences, the areas of the search space that will be traversed and the route each agent should take at each decision point in the activity

Modelling Distributed AI Systems as a Distributed Goal Search Problem

All coordination mechanisms can ultimately be reduced to commitments and their associated conventions (Jenning, 1993)

A. Commitments

Commitments can be seen to have a number of important properties. Agents can make pledges about both actions and beliefs and these pledges can either be about the future or the past.

B. Joint commitments

C. Conventions

D. Social conventions

Commitments provide the necessary structure for predictable interactions

Conventions provide the flexibility needed to operate in dynamic environments

Social conventions provide the necessary degree of mutual support

Coordination = Commitments + Conventions + Social Conventions + Local Reasoning