How Manifest Journals Transforms US Customs’ Import Data
We reveal the secrets behind the US Customs’ Import Data.
We introduce our name cracking technology.
We give the reasons why Manifest Journals is the choice of savvy users of the US Customs’ Import Data.
 
 
 
 
 
 
 
 
How We Apply a Clean US City, State and Zip Code to Each Bill of Lading

The raw US Customs’ Import Data doesn’t have a clean field for US city, state and zip code.

Free text entry into the Automated Manifest System (AMS) affects US importers’ addresses as much as free text entry affects US importers’ names.

You find two basic problems with US importers’ addresses in the raw data.

1. The parts of the US importers’ address-street name, city, state, and zip code- can be spread across 5 fields.

--Importer_Name
--Importer_Address_1
--Importer_Address_2
--Importer_Address_3
--Importer_Address_4

In this case we sift through the raw data’s 5 fields to find the text strings that match the US state, city and zip code.

2. Often the US state, city and/or zip code are missing from the record in the raw data.

Here we use the available data as clues to establish what the missing values ought to be.

We divide our process to assign a clean US state, city and zip code into 36 separate sub-routines. These 36 separate sub-routines work for 94% of the bills of lading.

We have to process the last 6% manually because the text in the raw data address fields are beyond the limit of a software routine.
Why We Need 36 Different Sub-Routines to Apply to US City, State and Zip Code

Here are some of the situations that we face:

1. We find the state, city and/or zip code is present in the raw data but the state, city and/or zip code is incorrect.

2. A zip code is present but a state and/or city is not present.

3. A city is present but a state and/or zip code is not present and the city has 1 or more state that could apply.

Here is the distribution of the 36 sub-routines:
 
Subroutine ID Bill Count %
1 1,762,840 50.2%
2 727,933 20.7%
3 178,314 5.1%
4 154,888 4.4%
5 138,193 3.9%
6 122,811 3.5%
7 77,640 2.2%
8 64,047 1.8%
9 58,963 1.7%
10 49,005 1.4%
11 46,113 1.3%
12-36 129,559 3.7%
 

Notes:

1. Percentages (%) are rounded up or down to the nearest tenth. The figures don’t add up to 100%.

2. Bill Count includes only house bills of lading where the address of the consignee is located in the United States.

We exclude from the count, consignees with foreign addresses. In Manifest Journals, you can still search consignees with foreign addresses.
How We Fix the Four Big Problems in the Raw Data from US Customs
 
 
 
 
 
 
 
 
 
 
 
Copyrights © 2009 Manifest Journals | All rights reserved Privacy Policy | Terms & Conditions About us| Contact us