Description
- Year: 2013
- Duration: 20 man hours
Introduction:
This research project was done to replace the human being at the reception at a
telecommunication company with an Artificial Intelligent agent.
Component:
1. Android application:
Through this application the clients can connect to the web service through mobile devices to communicate with the AI agent.
1. Android application:
Through this application the clients can connect to the web service through mobile devices to communicate with the AI agent.
2. ASP web service:
The devices that support android connect to this service.
The devices that support android connect to this service.
3. AI agent:
The AI agent responds to the clients by analyzing their questions.
4 .Web
application:The AI agent responds to the clients by analyzing their questions.
Through the web site also
the clients can connect to the AI agent.
Android application:
IMAGE
This application was developed using the IDE Eclipse,
Android 2.2 was used as the OS.
The user has to login to the application. The user’s
phone number will become the user name and it has a password also.
The main functions are the user can obtain a loan or
recharge the mobile using text.
The question will be passed to the web service and the
answers will be obtained through a jar file called ksoap2. The answers will be sent to the clients in
voice format.
Asp web service:
Developed in
visual studio 2010 with .net 3.5 frame work.
Here the
questions will be transferred to the business logic.
Business logic:
Here it will
be analyzed if the user needs a loan, or recharge the phone or user wants to
clarify any questions. Here it will be further analyzed by the AI agent if the
questions are general or specific to his own account.
Here the AI also uses spelling checker and similar word matching (word net).
Spelling checker example: the word “paymnt” will be analysed as payment.
similar words matcher: hello and hi will be considered as same words.
Here the AI also uses spelling checker and similar word matching (word net).
Spelling checker example: the word “paymnt” will be analysed as payment.
similar words matcher: hello and hi will be considered as same words.
AI agent
The AI agent was developed by using a markup language called AIML.
Following pdf is project SRS document
- Budget: Rs. 70,000
No comments:
Post a Comment