LSD: Large scale demonstrator
Large Scale Demonstrator
Large Scale Demonstrator Presentation (avi version)
Large Scale Demonstrator Presentation (mpeg version)
The Business-To-Business scenario involves a Central Business and several Client Businesses. Client Businesses engage in e-commerce operations via the Internet with the Central Business or among themselves. Typically, Client Businesses order goods from the Central Business. However, a Client Business can look for competitive pricing on a product among other Client Businesses, and may decide to purchase from the ‘best’ offer instead.
The interactions between businesses and the processing carried out within each business offer a variety of opportunities for applying argumentation-based processing techniques. Argumentation-based inference can be used by a business partner in order to deduce elements about the credit worthiness of a client. Argumentation-based decision-making is applicable when a business partner needs to decide whether to proceed with a transaction, whether this means accepting a request for a credit limit increase or accepting a purchase offer. Argumentation-based dialogue is a powerful tool for negotiating transaction terms. Finally, argumentation-based machine learning can help a business tune its assessment of one of its business partners based on its record of transactions. These interactions generated in the B2B context will also leverage argumentation-based scenarios for product negotiation and information seeking
The framework of the LDS is shown at the figure bellow.
System realization
This document covers the functionality of the Client Part (CP) of LSD system. The Server or Central Unit (CU) functionality is very like to Client and has only some additional modules for running under Apache server.
The CP is an Java application installed on a local computer or workstation. It provides a user with ability to realize B2B scenarios with other clients of the LSD system. The list of provided functionalities includes:
ü User management functionality
o Open ‘Access’ tab to see users’ information and manage them. It allows create, update and delete user accounts information, login, passwords, registration terms, rights and user’s profile.
ü Goods management functionality
o Open ‘Goods’ tab to see goods information and goods management. This information is stored in the local database.
ü Semantic rules management functionality
o Open ‘Rules Editor’ to see Business Rules information and rules management features. This part is realized with use of SQL-like queries that implements common and special business rules. User has ability to create own business rules using predefined system variables and rules/. Special editor is developed for this aim.
o Other approach uses Argumentation Engine module. Editor allows open current argumentation database, make test queries on it , modify and save changes. Rules is represented as prolog-like sentences.
ü Clients history review and analyzing functionality
o Open ‘Client History’ tab with Clients’ business history and business rules results. This part works with results stored in the local database. On this page some you can see results of analysis of client history with use of business rules. Results are presented at the bottom of the page.
o The similar page is realized for Argumentation database. Client history is analyzed by argumentation engine and present report about selected client.
ü Set of tools for application management
o Crush & Recovery
§ Open a window of backup and recovery service.
o Show user’s session
§ Show a log of the last user’s sessions
o Show user activity
§ Show a log of the user’s activity at the period of day, week, two weeks,total.
o Available Servers –
§ Show a list of available servers.
o Connect/Disconnect
§ Connect/disconnect with the Central server
o List of backup servers
§ Receive list of backup servers from the central server and open a tab where synchronization with these servers are allowed
o List of Clients
§ Receive a list of all clients from the Central server with their information (no editing is allowed now)
ü Semantic search
o Start/stop semantic search server allows client computer to be available for searching of goods from the other LSD application
o Search page allows make requests to known search server. The search uses mobile agents technology to perform connections and deliver information between clients and central server. Search engine uses TopicMaps semantic engine to fulfill and represent search results.
o Search results are delivered to the central server special page and can be viewed on the Search result page.
Last efforts were made to integrate Argumentation Engine with LSD application. Now we have made the next works.
- Simple standalone application with integrated Argumentation Engine was created
- Prolog base business rules for common and special business rules was developed
- Database with Client History data was connected with application.
- View and editing of business rules functionality were developed.
The application ArgViewer consists of two pages.
The first page contains a window for displaying of the loaded argumentation database. User can edit argumentation database, perform simple query and apply changes. All facts and rules should be written in prolog language.
The second page shows Client history data from the local database. Application takes user data and checks them against of the Argumentation database. Results of processing are shown to the user.
Below we place a diagram that shows main modules of the LSD application.
To enlarge the image press the figure 1: Figure 1
The next schema below is represented the general architecture of current version of the system with the argumentation component that is planned to add in the nearest future.
Argumentation plug-in in this schema is the module which will support argumentation dialogue. It will use Argumentation and Decision engines to communicate with argumentation database (zeus.arg). Client and server plug-in will interact by use of SOAP messages through TOMCAT.
The server and client components was realized for LSD system.
Server part is realized as Tomcat’s web application and communicate with Client part by use of SOAP. Semantic search components realize communication by TCP socket connections by port 6666/6667.
Architecture of LSD system
To enlarge the image press the figure 2: Figure 2
Client part is realized as separate GUI application that support SOAP messaging to communicate with server. It also supports work with argumentation database by use Argumentation Engine API.
Client and server both use MySQL database for storing data.
Explaining of information that can be discovered in application
The next part describes a set of information that user can view on separate pages of LSD application. This information is stored in the database and can make a general representation about application functionality.
Access tab
Allow to view and manage access information.
Columns:
Access - shows user login information. On click you may change login, password start and expire dates.
Expire Date – shows a date when login will be denied.
User Type - shows predefined user types. On click you may change it
User Right - shows 1/0 encrypted user rights (editing not realized yet)
User - show the first and the last user’s names. On click you may view and edit user’s detailed information.
Email - shows user e-mail address
Appointment – shows user’s appointment. On click you may change it.
Client - shows User’s company name. On click you may view detailed information and change it.
Goods tab
Allow to view and manage goods information.
Columns:
Name – shows a name of a good. On click you may view detailed information and change it.
Model – shows a model information of a good.
Version - shows a version of a good.
Price – shows a price of a good, in euro.
Publication date - shows a time when this good was published.
Keyword - shows a set of keywords that related to the good. On click you may change it.
Unit – shows a minimal amount of the goods in the trade unit. On click you may change it.
Producer – shows the good Producer. On click you may view producer’s list and change it.
Country – shows a country where the good was produced. On click you may change it.
Rules tab
Allow to view and manage business rules.
The left column contains a list of business rules of two types. Click on the rule will show its content on the right panel.
User can create custom rules and edit it with use of predefined system variables and existing rules.
Editor use if-then—else syntax to make the work more easy. Rules has a strict order in the left part and performed by engine from top to bottom.
Rules 2 tab
Allow to view and manage argumentation database.
User can see current argumentation database, make and apply changes. He also can execute prolog queries against database.
Rules are written in prolog-like language that is used by Argumentation Engine.
Client history tab
Allow to view client history and view results of business rules for each client. All names were taken from excel files with client history.
Columns:
CLN_ID - shows a client id in the system.
Client Name – shows a client name.
Business operates – shows a term of client’s company exists, in months.
Client for – shows a term of client work with the system, in months.
All other columns have clear names. Sums are in the thousands of euro
Credit given in this period(month)
Past credits
Payment made in this period (month)
Credit owned
Number of past debts paid before the due date
Number of past debts paid on the due date
Number of past debts paid after the due date
Average days past till pay after the due date
Previous period (moth) turnover
Last 6 moths turnover
Last 1 year turnover
Starting capital
Below the table you can see the table of implemented business rules to make a decision about possible credit limit changing.
User can also check his custom rule on selected client. He has a list of custom rules and can select and calculate result on it.
Client history 2 tab
Allow to view client history and view results of argumentation rules for each client.
The Table information is the same as in the previous tab. Recommendation obtained from the argumentation database id displayed at the bottom of page.
Client History new
This selection allows the user to view all the companies in the database and also to review and edit the argumentation rules which will be used by the underlying Inference Engine. There’s also a “REPORT CHN” button with which the user can enter the specific number of the companies the user chooses to test with the Inference Engine. When the test ends a new file is created under the name “REPORTCreditHistoryNew.csv” which is a report for each company tested.
Client History
This selection allows the user to test all the companies or a certain number of them against an existing set of SQL rules. The user can change this SQL rules set with the help of the View->Rules editor menu selection. The SQL rules procedure is the same as before and at the end of the tests a file under the name “REPORTCreditHistoryOld.csv” is created where the report for all companies tested exists.
