+ Project work flow and description (Business Profile extraction) at the very first time.

BusinessSearch Project work flow and description

Objective of BusinessSearch Project:

Main goal of this project is to collect the business profile in a automated way and try to develop a system that can update the profile in quarter or half year basis.

Flow Description:

The target of our project is to collect maximum business profile. But the business profile is fully depends on host name. If we have a single host name then we are able to create a profile by using it’s home page, contact us, about us pages. But we do not have enough host names to extract Business profile for our project . That is why our main target is to collect maximum host name and convert it to a perfect business profile.

Current / available attributes for Business profile are:

·         Fax

·         Email

·         Website

·         Zip code

·         Address

·         Phone no

·         GPS location

·         Contract person

·         Contact person’s designation

Upcoming/ future attributes are:

·         Business type

·         Business Description

·         Owner/ Proprietor/ Director

·         Establishment date

·         Registration Date

·         Logo

Challenges: WebCrawler

·         Robustness

·         Mirror

·         Hashing

·         Unexpected bug

Work done so far in Business Search:

·         Fax : 80- 90% accuracy

·         Phone: 80-90%

·         Zip Code: 70-80%

·         Website: 85-90%

·         Email: 85-90%

·         Address: 45-50%

·         Company Name: 45-50%

·         Contact Person: 50-60%

·         Contact Designation: 60-70%

·         GPS Cordinate: 60-70% but (we will consider it later with the support of iSearch/Lucene)

·         Branch: 70-80%

Yet to Develop:

·         Business type – (Machine Learning)

·         Business Description – (Snippet)

·         Owner/ Proprietor/ Director (Natural Language Processing)

·         Establishment date - Parsing

·         Registration Date – Parsing

·         Logo – Paring + Tricks + Crawling