Relational Database

0CONSUMER COMPLAINT DATABASE
Xuan Han – Da Jiang
Rohan Singla – Steven Spangler
CIS 8040
11/21/2016
CONSUMER COMPLAINT
DATABASE PROJECT

The Consumer Financial Protection Bureau (CFPB)1
is a federal agency that “helps consumer finance
markets work by making rules more effective, by consistently and fairly enforcing those rules, and by
empowering consumers to take more control over their economic lives.” One way the CFPB carries out this
stated mission is by providing a forum for disgruntled consumers to register their complaints against financial
service companies by whom they may have been wronged. The agency will then forward those complaints to
the offending company, allowing the company to right their wrong or respond with their side of the story.
Additionally, the CFPB makes much of this valuable information public through the upkeep an excellent data
set of all the complaints it receives from consumers. This data set is consistently updated and offers many
details about the complaints filed as well as the way companies respond to those complaints.
The members of this project team, as consumers of financial products ourselves, know that finding
objective information about the customer service performance of financial product providers is very difficult.
While the public is constantly exposed to the heavily curated marketing efforts of banks and lending institutions
large and small, there simply isn’t an easy way to predict exactly how satisfactory one’s experience might be as a
customer of these companies. And while the CFPB makes their information available to all, it doesn’t come in an
aggregated form that would make for a quick view of other consumers’ financial experiences.
Given this problem, our team sought to design and implement a database that would store all the
complaint information provided by the CFPB along with other demographic data that would combine with what’s
provided with each complaint to reveal objective insights. We wanted, for example, to be able to tally the number
of complaints against specific companies, see which companies have the most complaints registered against them,
and calculate rates of complaints for geographic and demographic portions of the country. And while the initial
questions we wanted to answer would be most useful to consumers, it was easy to see that we could also provide
insights that would be useful to financial service companies. They could benefit from knowing such facts as which
products gave consumers the most problems industrywide or whether the number of complaints correlated with
the average level of income in a geographic area.
The following sections of this report will cover the methodology used to create this database as well as the
answers and accompanying questions that querying the database provides.
We relied on two main sources for the data that would be inputted into the database –
 The full data file from the Consumer Complaint Database section of the CFPB website2
, available in
Comma Separated Value format
 Various tables (in CSV format) from the United States Census Bureau’s American Fact Finder website3
,
which provides a Guided Search function to assist in sifting through the mountains of available data
Additionally, we used individual Google searches to find contact information for some of the large public
companies referenced in the Consumer Complaint Data file.
1
http://www.consumerfinance.gov/
2
http://www.consumerfinance.gov/data-research/consumer-complaints/
3
http://factfinder.census.gov/faces/nav/jsf/pages/index.xhtml
Data Gathered
Objective

The data from the CFPB comes in a flat CSV file. There is one complaint per row, each assigned an
identification number, with additional attributes of the following types –
 What the complaint is about
o Product, Sub-product, Issue, Sub-issue, Company
 Who submitted it
o For privacy reasons, only the consumer’s State and Zip Code, as well as their method of
submission
 The company’s response
o When complaint was submitted to the company, whether the company responded in a timely
fashion, a categorized marker of the nature of the response including whether relief (monetary
or non—monetary) was provided, any public remarks the company wished to make, and
whether the consumer disputed the company’s response
Our team sought to add the following demographic information for the populations in the given consumer
states and zip codes –
 Population levels from the last census as well as updated estimates
 Educational attainment
 Average income level
Our team decided to treat each of those groups of complaint attributes as separate entities. Additionally,
though the CFPB must limit the amount of information they release about consumers who submit complaints, we
wished to model our database as if we had more information about the consumers. We thus chose to create a
consumer entity as well as assign each complaint to a unique consumer (with assigned identification number),
despite the fact that in reality each complaint doesn’t necessarily come from a unique consumer.
The eight entities we selected for our relational model were Complaints, Responses, Companies,
Products, SubProducts, Consumers, States, and Zip Codes. As just mentioned, we assume a one-to-one
relationship between Consumers and Complaints. Consumers are in a one-to-many relationship with both States
and Zip Codes – each consumer has one state and zip code, but each state and zip code has many consumers in
it. Each Complaint has one Response, and each Response is to one Complaint. A Company can be mentioned in
many Complaints, but each Complaint is in regard to one Company, and similarly between Responses and
Companies. Products and Subproducts are in a one-to-many relationship with Complaints – each complaint has
one Product and optionally one Subproduct, and Products and Subproducts appear in many Complaints.
The ER Diagram for the database from the MySQL Workbench software is on the following page.
Database Design

➢ Appendix II/ ER model and related table

The complaint data downloaded from the CFPB comes in a flat table in second normal form, with all the
non-key attributes determined by the ComplaintID. Because we wished to add more data, such as company
contact information, we began by splitting the table attributes into individual tables to match our designated ER
structure.
For reasons discussed previously, ConsumerIDs were added, and the functionally dependent State and
Zip Code values were sent to the Consumer table (as foreign keys referencing the State and Zip tables).
CompanyName was split away to the Company table, and since the values for CompanyName were unique and
uniform in the data, this was chosen to be the primary key for the table. Contact attributes were added as well.
And attributes pertaining to company responses were split off into the Response table. To avoid having a
composite primary key of both ComplaintID and CompanyName for this table, ResponseID was created as
primary key, with ComplaintID and CompanyName remaining as foreign keys.
We chose to assign ID numbers to both the 12 products and 48 subproducts present in the complaint
data for several reasons. The string values for these attributes can be lengthy, so an ID number prevents entry
errors and mismatches. Also, providing the names in a separate table allows for easy future additions and
modifications to the range of products/subproducts and prevents deletion anomalies in the Complaint table.
One thing our team noticed was that many of the zip codes reported for complaints only contained the
first three digits followed by an ‘XX’ string. Because of this (and also in order to avoid ‘overgranualizing’ our
geographic data), we realized we would need to group zip codes by their first three digits, which correspond to
distinct geographic areas. This would become tedious when scripting queries, so we elected to create and store a
‘Zip3’ dependent attribute for both the Zips and Consumer tables. Storing this value would improve performance
and ease query writing.
No further normalization was required, as this structure meant each table was in third normal form, free of
partial and transitive dependencies. We have provided the structure of our related tables on the following page.
The script written to create the physical tables of the database can be found in Appendix II at the end of
this report. Data types for each attribute were selected based on a quick look through the possible values in the
CSV files as well as the descriptions given on the CFPB website for each piece of data they share4
.
The flat CSV complaint file was split into separate CSV files, one for each table. The CSV files from the
American Fact Finder website required some cleaning to remove extraneous data, but similarly existed with one
for each table, Zips and States. The ‘LOAD DATA INFILE’ command was used to enter the data from the CSV
files. Our team learned that this command is very efficient at quickly loading data into a MySQL database from
CSV files, making quick work of 600,000+ rows of complaints.
4
http://www.consumerfinance.gov/complaint/data-use/
Logical Tables &
Normalization
Table Creation
& Data Entry

Responses
ResponseIDComplaintIDCompanyNameResponse_to_consumerTimelyCompanyPublicResponseConsumerDisputed
Products
ProductIDProductName
SubProducts
SubProductIDSubProductName
Complaints
ComplaintIDCompanyNameConsumerIDSent_to_company_DateRec_DateHow_SubmittedProductIDSubproductIDIssueSubIssue
Consumer
ConsumerIDStateZipCodeZip3
Company
CompanyNamecomphonecomstreetcomcitycomstatecomzip
States
StateAbbStateFullNameStateAreaStateMedianEarningsStateHSorHigherStateBachelorsorHigherStateGraduateDegStatePop25overStatePop2010StatePop2011StatePop2012StatePop2013StatePop2014StatePop2015
Zips
ZipCodeZipAreaZipMedianEarningsZipHSorHigherZipBachelorsorHigherZipGraduateDegZipPop25overZipPop2010ZipPop2011ZipPop2012ZipPop2013ZipPop2014ZipPop2015Zip3

In this section, we pose the questions for which the constructed database can provide answers.
Queries, outputs, and some additional analysis are included.
Which states have the highest and lowest rate of complaints?
1. SELECT a.StateAbb,
2. (a.TotComplaint/b.statepop25over)*1000
3. as ComplaintsPer1000
4. FROM
5. (SELECT cn.StateAbb,
6. count(cm.complaintid) as TotComplaint
7. FROM complaints cm, consumer cn
8. WHERE cm.consumerid = cn.consumerid
9. GROUP BY cn.stateabb) a
10. LEFT JOIN
11. (SELECT StatePop25over, StateAbb FROM states) b
12. ON a.StateAbb = b.StateAbb
13. ORDER BY ComplaintsPer1000 DESC LIMIT 10;
1. SELECT a.StateAbb,
2. (a.TotComplaint/b.statepop25over)*1000
3. as ComplaintsPer1000
4. FROM
5. (SELECT cn.StateAbb,
6. count(cm.complaintid) as TotComplaint
8. WHERE cm.consumerid = cn.consumerid
9. GROUP BY cn.stateabb) a
10. LEFT JOIN
11. (SELECT StatePop25over, StateAbb FROM states) b
12. ON a.StateAbb = b.StateAbb
13. ORDER BY ComplaintsPer1000 LIMIT 10;
We have queried for the total number of complaints in each state divided by the population of adults 25 and over.
Multiplying by 1,000 scales the result to a digestible number.
The District of Columbia has by far the highest rate of complaints against financial companies. It’s neighbors,
Delaware and Maryland come next. Including Virginia, we see that the Mid-Atlantic region proves to be
problematic for consumers’ interactions with financial products.
The states with the lowest rate of complaints are mostly in the upper Midwest and south along the Mississippi
River, as well as Appalachian neighbors Kentucky and West Virginia. These results seem fairly regionalized as
well.
Queries

Which companies received the most complaints overall?
1. SELECT count(cpl.ComplaintID) as NumComplaints,
2. cpl.CompanyName, cpy.ComPhone, cpy.ComStreet,
3. cpy.ComCity, cpy.ComState, cpy.ComZip
4. FROM complaints cpl, company cpy
5. WHERE cpl.CompanyName = cpy.CompanyName
6. GROUP BY cpl.CompanyName
7. ORDER BY NumComplaints DESC
8. LIMIT 30;
These are the top 30 companies in number of complaints received. Our database has contact information for
these companies, which make up a large majority of all complaints received.
Just how many of the total complaints are lodged against the top 30?
1. SELECT sum(a.NumComplaints)/b.Tot as 'Percentage of All Complaints'
2. FROM
3. (SELECT count(cpl.ComplaintID) as NumComplaints, cpl.CompanyName, cpy.ComPhone,
4. cpy.ComStreet, cpy.ComCity, cpy.ComState, cpy.ComZip
5. FROM complaints cpl, company cpy
6. WHERE cpl.CompanyName = cpy.CompanyName
7. GROUP BY cpl.CompanyName ORDER BY NumComplaints DESC LIMIT 29) a,
8. (SELECT count(ComplaintID) as Tot FROM complaints) b;
NumComplaints CompanyName ComPhone ComStreet ComCity ComState ComZip
60560 Bank of America 704-386-5681 100 North Tryon Street Charlotte NC 28255
47561 Wells Fargo & Company 866-249-3302 420 Montgomery St. San Francisco CA 94104
40719 Equifax 404-885-8000 1550 Peachtree Street NE Atlanta GA 30309
38547 Experian 1 714 830 7000 475 Anton Blvd. Costa Mesa CA 92626
38111 JPMorgan Chase & Co. 1-212-270-6000 270 Park Ave New York NY 10017
32785 TransUnion Intermediate Holdings, Inc. 312-985-2000 555 West Adams Street Chicago IL 60661
30442 Citibank 1-800-285-3000 399 Park Ave. New York NY 10043
22669 Ocwen (800) 746-2936 12001 Science Drive, Ste. 110 Orlando FL 32826
17701 Capital One 1-877-383-4802 1680 Capital One Drive McLean VA 22102
14772 Nationstar Mortgage 1-888-708-4043 8950 Cypress Waters Blvd. Coppell TX 75019
11027 Synchrony Financial 1 (866) 226-5638 200 Crossing Blvd Suite 101 Bridgewater NJ 08807
10894 U.S. Bancorp 1.800.934.6802 111 Fillmore Avenue East St Paul MN 55107
10302 Ditech Financial LLC (800) 700-9212 1100 Virginia Dr #100a Fort Washington PA 19034
10005 Navient Solutions, Inc. (703) 810-3000 2001 Edmund Halley Dr Reston VA 20191
7789 PNC Bank N.A. (888) 762-2265 222 Delaware Avenue Wilmington DE 19899
7077 Amex 1 (800) 528-4800 200 Vesey Street. New York NY 10285
6981 Encore Capital Group (877) 445-4581 3111 Camino Del Rio N #103 San Diego CA 92108
6668 HSBC North America Holdings Inc. 212 525 5000 452 5th Avenue New York NY 10018
5555 Discover 1-800-347-2683 502 E. Market Street Greenwood DE 19950
5515 SunTrust Banks, Inc. (404) 588-7711 303 Peachtree Street Northeast Atlanta GA 30308
5220 TD Bank US Holding Company 888-751-9000 1701 Route 70 East Cherry Hill NJ 08034
5082 Select Portfolio Servicing, Inc (801) 293-1883 3217 Decker Lake Dr Salt Lake City UT 84119
4162 Portfolio Recovery Associates, Inc. (731) 984-7680 4829 US-45 Jackson TN 38305
3563 Citizens Financial Group, Inc. (401) 456-7000 1 Citizens Plaza Providence RI 02903
3392 Barclays PLC 1 212 526 7000 745 7th Avenue New York NY 10019
3304 Seterus, Inc. (866) 570-5277 14523 SW Millikan Way Beaverton OR 97005
3251 Fifth Third Financial Corporation (800) 972-3030 38 Fountain Square Plaza Cincinnati OH 45263
3170 ERC (256) 430-3080 308 Voyager Way Huntsville AL 35806
2951 BB&T Financial 800-226-5228 200 West Second Street Winston-Salem NC 27101

Does the level of educational attainment for an area correlate with the number of complaints
from that area?
1. select co.Zip3, zi.Bach/zi.Zip3Pop as 'Bach+',
2. zi.Grad/zi.Zip3Pop as 'Grad+',
3. zi.NoHS/zi.zip3pop as LessThanHS,
4. count(c.complaintid) as TotalComplaintsReceived
5. from complaints c, consumer co, (SELECT Zip3,
6. sum(ZipPop25over) as Zip3Pop,
7. sum(zipbachelorsorhigher*zippop25over) as Bach,
8. sum(zipgraduatedeg*zippop25over) as Grad,
9. sum((100-zipHSorHigher)*zippop25over) as NoHS
10. FROM zips GROUP BY zip3) zi
11. where c.consumerid=co.consumerid and co.zip3 = zi.zip3
12. group by co.zip3
13. order by TotalComplaintsReceived desc;
The query returns weighted averages for the percentage of people in each of the 893 3-digit zip code areas who
have
 Earned at least a bachelor degree
 Earned a graduate degree
 Not completed high school
Further analysis is required to visualize the results of this query. Results follow on the next page.

Graphs represent the distribution of total
complaints against the average level of
education in each 3-digit zip code area.
The top and bottom charts show a
distribution skewed to the right, with
more complaints on the side of lower
percentage of educational attainment. It
can be seen that more complaints have
been received from areas of lower
education attainment.
 Areas where percentage of people
having bachelor degree or higher
between 16 to 45 percent have
registered higher number of
complaints
having graduate degree is between
5 to 25 percent have registered
higher number of complaints
who haven’t completed high
school is between 5 to 20 percent
have registered higher number of
complaints.

Does the average income of an area correlate with the number of complaints made from it?
1. SELECT co.zip3,
2. zi.TotEarn/zi.TotPop as 'Median Income',
3. COUNT(c.complaintid) AS TotalComplaintsReceived
4. FROM
5. complaints c, consumer co,
6. (SELECT Zip3,
7. sum(ZipMedianEarnings*ZipPop25over) AS TotEarn,
8. sum(ZipPop25over) as TotPop
9. FROM zips GROUP BY zip3) zi
10. WHERE c.consumerid = co.consumerid AND co.zip3 = zi.zip3
11. GROUP BY co.zip3
12. ORDER BY TotalComplaintsReceived DESC;
The query has taken a population-weighted average of median income in each zip code of all 893 3-digit zip code
areas, and has also returned the total number of complaints from each 3-digit zip area. We will again have to
provide further analysis to reveal useful insight.
Most complaints in the database come from areas with median annual income of $25-$35k. The
distribution has slight right skewness, indicating that more complaints come from the lower income
side.

Do certain companies have more complaints in areas with higher or lower population density?
The database includes the area of each zip code, so we can calculate population density. The query returns the company
with the most complaints in each 3-digit zip code area, and is then ordered by that area’s population density.
1. SELECT a.Zip3, z.Density, a.CompanyName, a.NumComplaints
2. FROM (SELECT DISTINCT Zip3, CompanyName,
3. count(cm.CompanyName) as NumComplaints
5. WHERE cm.consumerID = cn.consumerID
6. GROUP BY cn.Zip3, cm.CompanyName
7. ORDER BY NumComplaints DESC) as a,
8. (SELECT Zip3, sum(ZipPop2014)/sum(ZipArea) as Density
9. FROM zips GROUP BY Zip3) as z
10. WHERE a.Zip3 = z.Zip3
11. GROUP BY a.Zip3
12. ORDER BY z.Density DESC, a.NumComplaints DESC
13. LIMIT 15;
The top 3-digit zip areas in the US are mostly New York City and its
boroughs, which are those beginning with 1 in this table as well as those
beginning with 07 (New Jersey). We also see 941 (San Francisco), 903
(Los Angeles), 022 (Boston), and 6060 (Chicago). The megabanks easily
dominate the complaints in these densely populated areas.
1. SELECT a.Zip3, z.Density, a.CompanyName, a.NumComplaints
2. FROM (SELECT DISTINCT Zip3, CompanyName,
3. count(cm.CompanyName) as NumComplaints
5. WHERE cm.consumerID = cn.consumerID
6. GROUP BY cn.Zip3, cm.CompanyName
7. ORDER BY NumComplaints DESC) as a,
8. (SELECT Zip3, sum(ZipPop2014)/sum(ZipArea) as Density
9. FROM zips GROUP BY Zip3) as z
10. WHERE a.Zip3 = z.Zip3
11. AND z.Density != 0
12. AND a.NumComplaints >= 10
13. GROUP BY a.Zip3
14. ORDER BY z.Density ASC, a.NumComplaints DESC
15. LIMIT 15;
This list of sparsely populated US 3-digit zip areas doesn’t seem to reveal
any particularly great insights. The companies receiving the most
complaints in these areas mostly mirror the list of companies receiving the
most complaints overall. Many of these areas are in the western portion of
the country, and thus we see Wells Fargo earning more complaints. Areas
represented in this table are Alaska (996/997/995), Elko, Nevada (898),
Great Falls, Montana (594), and Sheridan, Wyoming (828).

Do certain products have more complaints than others?
1. SELECT b.ProductName, a.TC as 'Total Complaints'
2. FROM (SELECT c.ProductID, count(c.ProductID) as TC
3. FROM complaints c
4. GROUP BY c.productID) as a
5. INNER JOIN
6. (SELECT p.ProductID, p.ProductName FROM products p) as b
7. ON a.ProductID = b.ProductID
8. ORDER BY a.TC DESC;
Mortgages have garnered the highest number of complaints, followed by Debt Collection and Credit Reporting.
Considering the difficulties presented by payday loans, such as high interest rates and predatory business practices,
one might be surprised to find the relatively small number of complaints lodged against this product.
Which companies have the highest number of complaints for specific products appearing most frequently in the database?
1. SELECT c.CompanyName, p.ProductName, count(c.ProductID) as NumComplaints
2. FROM complaints c, products p
3.
4. WHERE c.ProductID=p.ProductID
5. AND p.ProductName='Mortgage'
6. GROUP BY c.CompanyName
8. LIMIT 5;
4. AND p.ProductName='Debt collection'
7. LIMIT 5;
4. AND p.ProductName='Credit reporting'
7. LIMIT 5;

How do rates of timely response to complaints differ among the companies?
1. SELECT a.CompanyName, b.YesCount/a.Total as TimelyRate
2. FROM
3. (SELECT CompanyName, count(ComplaintID) as Total
4. FROM responses
5. GROUP BY companyname
6. HAVING count(ComplaintID) > 100) a,
7. (SELECT CompanyName, count(ComplaintID) as YesCount
8. FROM responses WHERE Timely = 'Yes'
9. GROUP BY CompanyName
10. HAVING count(ComplaintID) > 100) b
11. WHERE a.CompanyName = b.CompanyName
12. ORDER BY TimelyRate
13. LIMIT 25;
Of those companies that have at least 100 complaints against them in the
database, this is the list of those with the worst rate of timely responses.
Consumers should beware not receiving any response if they file a
complaint against one of these companies.
1. SELECT count(a.CompanyName)
2. FROM
4. FROM responses
12. AND b.YesCount/a.Total = 1;
On the other hand, of those companies with at least 100 complaints, 70
have a perfect record of timely responses.
1. SELECT a.CompanyName, b.YesCount/a.Total as TimelyRate
2. FROM
4. FROM responses
12. ORDER BY TimelyRate DESC
13. LIMIT 25;
The best performers from those companies appearing frequently in the
data (at least 500 complaints) appear to the right. All these companies
have a 99.9% response rate or better. Good job! We see the three
major credit bureaus here as well as some large banks and lenders like
BBVA, SunTrust, Barclays, and Sallie Mae.

How does the rate of consumer dispute of company responses differ among companies?
1. SELECT a.CompanyName, b.NoCount/a.Total as NotDisputedRate
2. FROM
4. FROM responses
7. (SELECT CompanyName, count(ComplaintID) as NoCount
8. FROM responses WHERE ConsumerDisputed = 'No'
12. ORDER BY NotDisputedRate DESC
13. LIMIT 25;
Of the companies who have at least 100 complaints in the database, we
have the list of those with the best rates of avoiding dispute from the
consumers they respond to. Consumers are happiest with the
responses they receive from this list of companies.
1. SELECT a.CompanyName, b.NoCount/a.Total as NotDisputedRate
2. FROM
4. FROM responses
7. (SELECT CompanyName, count(ComplaintID) as NoCount
8. FROM responses WHERE ConsumerDisputed = 'No'
12. ORDER BY NotDisputedRate
13. LIMIT 25;
On the other end of the spectrum, we have the rates of dispute of
larger companies that appear at least 500 times in the data, with the
poorest rate first. These companies often receive dispute to their
responses.

What are the most common responses by companies to complaints?
1. SELECT r1.Response_to_consumer,
2. count(r1.ResponseID) as NumResponses,
3. count(r1.ResponseID)/r3.Tot as Percentage
4. FROM responses r1,
5. (SELECT count(ResponseID) as Tot
6. FROM responses r2) r3
7. GROUP BY Response_to_consumer
8. ORDER BY NumResponses DESC;
Almost three-quarters of consumers are simply given
an explanation in response to their complaint. Only
about one in five are closed with relief, monetary or
non-monetary, provided.
What are the types of issues consumers complain about for each product?
1. SELECT p.ProductName, c.Issue, count(c.ComplaintID) as NumComplaints
4. GROUP BY c.ProductID, c.Issue
5. ORDER BY p.ProductName, NumComplaints DESC;
This query returns the complete list of Issues and the Product each issue corresponded with on the complaints. A
formatted table of the results is below.
ProductName Issue NumComplaints ProductName Issue NumComplaints ProductName Issue NumComplaints
Credit card Billing disputes 12803 Bank account or service Account opening, closing, or management 32056 Money transfers Fraud or scam 1621
Credit card Other 7853 Bank account or service Deposits and withdrawals 19626 Money transfers Other transaction issues 1228
Credit card Identity theft / Fraud / Embezzlement 7159 Bank account or service Problems caused by my funds being low 10382 Money transfers Money was not available when promised 798
Credit card Closing/Cancelling account 5492 Bank account or service Making/receiving payments, sending money 6211 Money transfers Other service issues 415
Credit card APR or interest rate 5029 Bank account or service Using a debit or ATM card 5141 Money transfers Wrong amount charged or received 221
Credit card Late fee 3217 Consumer Loan Managing the loan or lease 12379 Money transfers Incorrect/missing disclosures or info 169
Credit card Customer service / Customer relations 2995 Consumer Loan Problems when you are unable to pay 6294 Mortgage Loan modification,collection,foreclosure 104726
Credit card Delinquent account 2734 Consumer Loan Taking out the loan or lease 3541 Mortgage Loan servicing, payments, escrow account 68870
Credit card Credit determination 2672 Consumer Loan Shopping for a loan or lease 1702 Mortgage Application, originator, mortgage broker 15239
Credit card Advertising and marketing 2574 Consumer Loan Managing the line of credit 709 Mortgage Settlement process and costs 7923
Credit card Rewards 2476 Consumer Loan Account terms and changes 464 Mortgage Credit decision / Underwriting 5040
Credit card Credit card protection / Debt protection 2321 Consumer Loan Shopping for a line of credit 277 Mortgage Other 4866
Credit card Billing statement 2284 Consumer Loan Charged fees or interest I didn't expect 245 Other financial service Fraud or scam 252
Credit card Transaction issue 2259 Consumer Loan Can't contact lender 66 Other financial service Customer service/Customer relations 212
Credit card Payoff process 2027 Consumer Loan Lender repossessed or sold the vehicle 65 Other financial service Unexpected/Other fees 83
Credit card Credit line increase/decrease 1933 Consumer Loan Payment to acct not credited 59 Other financial service Excessive fees 79
Credit card Other fee 1909 Consumer Loan Applied for loan/did not receive money 26 Other financial service Disclosures 35
Credit card Credit reporting 1669 Consumer Loan Received a loan I didn't apply for 21 Other financial service Lost or stolen check 34
Credit card Unsolicited issuance of credit card 1471 Consumer Loan Can't stop charges to bank account 7 Other financial service Lost or stolen money order 32
Credit card Collection practices 991 Consumer Loan Lender damaged or destroyed vehicle 6 Other financial service Advertising and marketing 25
Credit card Balance transfer 982 Consumer Loan Charged bank acct wrong day or amt 5 Other financial service Incorrect exchange rate 15
Credit card Collection debt dispute 884 Consumer Loan Lender sold the property 4 Payday loan Charged fees or interest I didn't expect 2103
Credit card Forbearance / Workout plans 491 Consumer Loan Lender damaged or destroyed property 1 Payday loan Can't contact lender 836
Credit card Privacy 435 Credit reporting Incorrect information on credit report 84618 Payday loan Received a loan I didn't apply for 493
Credit card Application processing delay 433 Credit reporting Credit reporting company's investigation 13628 Payday loan Can't stop charges to bank account 421
Credit card Bankruptcy 410 Credit reporting Unable to get credit report/credit score 9519 Payday loan Payment to acct not credited 309
Credit card Sale of account 299 Credit reporting Improper use of my credit report 4623 Payday loan Applied for loan/did not receive money 283
Credit card Arbitration 298 Credit reporting Credit monitoring or identity protection 3732 Payday loan Charged bank acct wrong day or amt 240
Credit card Cash advance 220 Debt collection Cont'd attempts collect debt not owed 50370 Prepaid card Unauthorized transactions/trans. issues 1008
Credit card Overlimit fee 196 Debt collection Disclosure verification of debt 23871 Prepaid card Managing, opening, or closing account 1003
Credit card Balance transfer fee 192 Debt collection Communication tactics 20936 Prepaid card Fraud or scam 591
Credit card Cash advance fee 177 Debt collection False statements or representation 9859 Prepaid card Fees 197
Credit card Convenience checks 129 Debt collection Improper contact or sharing of info 8668 Prepaid card Adding money 177
Student loan Dealing with my lender or servicer 9839 Debt collection Taking/threatening an illegal action 7694 Prepaid card Advertising, marketing or disclosures 63
Student loan Can't repay my loan 5184 Virtual currency Other service issues 5 Prepaid card Overdraft, savings or rewards features 45
Student loan Repaying your loan 3786 Virtual currency Fraud or scam 4 Prepaid card Other 1
Student loan Problems when you are unable to pay 1676 Virtual currency Other transaction issues 2
Student loan Getting a loan 572 Virtual currency Money was not available when promised 1

Consider a fictional consumer who lives in Atlanta (30303) and is interested in acquiring a
checking account from a bank. What companies have received the most complaints in regards
to checking accounts in her area?
1. SELECT cm.CompanyName, count(cm.ComplaintID) as NumComplaints
3. WHERE cm.consumerID=cn.consumerID
4. AND cm.SubProductID='80880'
5. AND cn.Zip3=303
6. GROUP BY cm.CompanyName
7. ORDER BY NumComplaints DESC;
In order to make a decision on what institutions to avoid, this list of the
companies that have received the most complaints in the same 3-digit zip
area would be useful. There are also a handful of companies with one
complaint returned by this query.
What is the preferred method of submitting complaints to the CFPB?
The large majority of consumers prefer to use the web-based
form to submit their complaints.

Though our main concern in this project was not with modeling the entry of new data into the database,
the Consumer Finance Protection Bureau has a clear procedure. The agency makes available web, telephone, and
postal based interfaces (as seen from the query on the previous page) for the public to submit their complaints.
We have created the context-level and level-0 data flow diagrams below to model the system through which the
CFPB obtains and processes new complaints. Their processing involves dealings with the company who the
complaint is against (“Company” external entity below) and analysis from which the agency submits reports to
other government entities.
Appendix I - Data Flow

1. CREATE TABLE products(ProductName VARCHAR(40),
2. ProductID INT(2),
3. PRIMARY KEY (ProductID));
4. LOAD DATA LOCAL INFILE 'C:/Users/Steven Spangler/OneDrive - Georgia State University/CIS 8040/Project/finaltables/product.csv'
5. INTO TABLE products
6. FIELDS TERMINATED BY ',' OPTIONALLY ENCLOSED BY '"'
7. LINES TERMINATED BY 'n'
8. IGNORE 1 LINES;
9.
10. CREATE TABLE subproducts(SubProductName VARCHAR(40),
11. SubProductID VARCHAR(5),
12. PRIMARY KEY (SubProductID));
13. LOAD DATA LOCAL INFILE 'C:/Users/Steven Spangler/OneDrive -
Georgia State University/CIS 8040/Project/finaltables/subproduct.csv'
14. INTO TABLE subproducts
17. IGNORE 1 LINES;
18.
19. CREATE TABLE company(CompanyName VARCHAR(60) PRIMARY KEY,
20. comphone VARCHAR(16),
21. comstreet VARCHAR(50),
22. comcity VARCHAR(40),
23. comstate VARCHAR(2),
24. comzip VARCHAR(5));
25. LOAD DATA LOCAL INFILE 'C:/Users/Steven Spangler/OneDrive - Georgia State University/CIS 8040/Project/finaltables/company.csv'
26. INTO TABLE company
29. IGNORE 1 lines
30. (CompanyName,comphone,comstreet,comcity,comstate,@comzip)
31. SET comzip = nullif(@comzip,00000);
32.
33. CREATE TABLE States(StateName VARCHAR(20),
34. StateAbb VARCHAR(2) PRIMARY KEY,
35. StateArea DECIMAL(9,2),
36. StatePop25over INT,
37. StateGraduateDeg DECIMAL (3,1),
38. StateHSorHigher DECIMAL(3,1),
39. StateBachelorsorHigher DECIMAL (3,1),
40. StateMedianEarnings INT,
41. StatePop2010 INT,
46. StatePop2015 INT);
47. LOAD DATA LOCAL INFILE 'C:/Users/Steven Spangler/OneDrive - Georgia State University/CIS 8040/Project/finaltables/states.csv'
48. INTO TABLE states
51. IGNORE 1 LINES;
52.
53. CREATE TABLE zips(ZipCode VARCHAR(5) PRIMARY KEY,
54. ZipArea DECIMAL(7,3),
55. ZipPop2010 INT,
56. ZipPop2011 INT,
57. ZipPop2012 INT,
Appendix II - Data Entry Script

58. ZipPop2013 INT,
59. ZipPop2014 INT,
60. ZipPop25over INT,
61. ZipGraduateDeg DECIMAL (4,1),
62. ZipHSorHigher DECIMAL (4,1),
63. ZipBachelorsorHigher DECIMAL (4,1),
64. ZipMedianEarnings INT);
65. LOAD DATA LOCAL INFILE 'C:/Users/Steven Spangler/OneDrive - Georgia State University/CIS 8040/Project/finaltables/zips.csv'
66. INTO TABLE zips
69. IGNORE 1 LINES;
70.
71. CREATE TABLE consumer(StateAbb VARCHAR(2) REFERENCES states,
72. ZipCode VARCHAR(5) REFERENCES zips,
73. ConsumerID VARCHAR(6) PRIMARY KEY);
74. LOAD DATA LOCAL INFILE 'C:/Users/Steven Spangler/OneDrive - Georgia State University/CIS 8040/Project/finaltables/consumers.csv'
75. INTO TABLE consumer
78. IGNORE 1 LINES;
79.
80. CREATE TABLE complaints(Rec_Date DATE,
81. ProductID VARCHAR(5) REFERENCES products,
82. SubProductID VARCHAR(5) REFERENCES subproducts,
83. Issue VARCHAR(60),
84. Subissue VARCHAR(60),
85. CompanyName VARCHAR(60) REFERENCES company,
86. How_submitted VARCHAR(15),
87. Sent_to_company DATE,
88. ComplaintID VARCHAR(7) PRIMARY KEY,
89. ConsumerID VARCHAR(6) REFERENCES consumer
90. ResponseID INT(7) UNSIGNED PRIMARY KEY AUTO_INCREMENT);
91. LOAD DATA LOCAL INFILE 'C:/Users/Steven Spangler/OneDrive -
Georgia State University/CIS 8040/Project/finaltables/complaints.csv'
92. INTO TABLE complaints
95. IGNORE 1 LINES
96. (@Rec_Date, ProductID, SubProductID, Issue, Subissue, CompanyName, How_submitted, @Sent_to_company, ComplaintID, ConsumerID)
97. SET Rec_Date = str_to_date(@Rec_Date,'%c/%e/%Y'),
98. Sent_to_company = str_to_date(@Sent_to_company,'%c/%e/%Y');
99.
100. CREATE TABLE responses(CompanyName VARCHAR(60) REFERENCES company,
101. PublicResponse VARCHAR(130),
102. Response_to_consumer VARCHAR(40),
103. Timely VARCHAR(3),
104. ConsumerDisputed VARCHAR(4),
105. ComplaintID VARCHAR(7) REFERENCES complaints);
106. LOAD DATA LOCAL INFILE 'C:/Users/Steven Spangler/OneDrive - Georgia State University/CIS 8040/Project/finaltables/responses.csv'
107. INTO TABLE responses
110. IGNORE 1 LINES;
111.
112. ALTER TABLE `complaints`.`zips`
113. ADD COLUMN `Zip3` VARCHAR(3) NULL AFTER `ZipMedianEarnings`;
114. UPDATE zips SET Zip3 = left(ZipCode,3);
115.
116. ALTER TABLE `complaints`.`consumer`
117. ADD COLUMN `Zip3` VARCHAR(3) NULL;
118. UPDATE consumer SET Zip3 = left(ZipCode,

Relational Database

More Related Content

Viewers also liked (9)

Similar to Relational Database (20)

Relational Database