Implementation of Information Retrieval (IR) in

an Electronic Commerce Architecture using Back

Propagation Network Learning (BPNL) Algorithm

RIKTESH SRIVASTAVA, PHD

What the Research is all about?

Amazing Factors-Electronic Commerce

Trend is
decreasi
ng

Why the trend is decreasing?

Slow performance of the E-Commerce Architecture
12.32%
People are unable to find suitable E-Commerce locations
37%
People unable to find suitable products on E-Commerce
locations-51.63%

People unable to find suitable products on E-Commerce

locations-51.63% - Is there any solution?

Finding
Suitable
Products

Traditiona
l Solution

AI
Solution
(ANN)

Electronic Commerce Architecture

B2C EC

B2C-ANN Architecture

ANN Architecture-In
Detail

ANN Architecture-In
Detail
At this juncture, the
requests needs to be
searched from DB Server

Uses some weight for the

requests received at the
Input Layer

Requests
Received at
Input Layer

Additional Weights from

Hidden and Output Layer
and Process the
Information

Hidden
Layer

Output
Layer

Repetition
Activation
Function

BPNL Algorithm
Initialize weights for the request type (small random numbers)
For each training of user requests
Repeat until weights convergence or till a required number of epochs
are completed
Receive requests as it will be extracted from various queues
Propagate the error backward from output layer to hidden and input layer.
Calculate new weights in accordance with BPNL algorithm.

Replace old weights new weights as taken from training algorithm

After every t time units
Measure performance of each requests
Repeat until performance falls below a threshold level() else go to Step iv.
Set activation of input unit. Inputs to input layer will be actual packets that are to be
scheduled.
Compute output of hidden and output layer using sigmoid activation function
Output will be fed to weight decider module which will calculate the required change in
weights of the queues.

Research Outcomes

Results

Recommendations
Implementation of BPNL Algorithm at the DB server level
increases the performance of the system
Decreases the Response time of the system
Aims to implement the code for a test B2C EC Architecture
and evaluate the performance for other EC Architecture as
well.