Analysis and Prediction

As we approached the half way mark for our project, we would like to present a recap of what have been accomplished so far. We have been making more analysis that may help us redesign the 211 call tree to minimize call wait time. One major impact on call wait time is the type of calls that 211 receives. We have identified the most consuming type and the frequency of those calls, and then compared it to other type of calls.

  • Most time consuming calls are less frequent
  • Frequent calls take less time
  • The average call time ratio of the frequent to infrequent calls is about 1:10

This study provides great insight to fix the call menu and give the 211 team more options to better manage their staff.

2-1-1 center has a great deal of useful information that can be very valuable for organization and service providers to deliver better assistance to people in needs. For example, 2-1-1 records caller’s zip code along with caller’s needs. By comparing these types of information one can understand which type of services that are highly demanded in diverse areas; thus provide suggestions to service providing parties to consider extending their service coverage to the appropriate locations to more effectively service location in greater needs. Below is a heat map that shows a comparison between types of needs and different regions



As a work in progress we are also developing a predictive model for future call volume and total talk time based on several factors, such factors incorporated weather, week number, day of the week and GDP.

A preliminary result is shown below


For next week we will be exploring more factors, such as gas prices to get better prediction for future call volume and total talk time.


The Power of Analysis

Building relational database and the ability to able to use MySQL for multiple queries to do more data analysis was a major goal for our team this week. The process of building MySQL database included:

  • Extracting useful columns (ignoring mostly empty columns).
  • Creating table, columns to match Refer-Net column data.
  • Adding appropriate indices (category, actionservice, datetime fields, transitioned).
  • Verification that data made sense.
  • Creating data import tool to map Refer-Net exported data from XML to SQL- DB to push 2.5 years of data.

Our main goal is to identify problem areas, peaks and drops in call time. We are basing our assumption that by identifying these issues will contribute in re-constructing a better menu and thus decrease callers wait time; for more information on the 211 call center, check out last week’s blog post!

Using MySql database, we were able to perform more analysis; such an example is the analysis we have done on calculating the numbers of calls based on action-services Terms (Needs per call). In order to be more specific, we have completed some keyword clustering and were able to group them into smaller subsets, these subsets are shown in the pie chart below. The result is a little ambiguous because it shows that around 10% that is considered to be under the subset (Others). This needs more investigation and will require finding a good method to improve the result.



We also examined the abandoned calls data more carefully for specific months of the years 2014 & 2015. The following results are for February 2015.

Plotting the frequency of abandoned calls vs. time duration indicates the existence of outliers in our data which needs to be detected and removed.



The following plot shows the same results but with outliers removed from the dataset. Time is in seconds.


As shown above, most calls were abandoned during the first 10 seconds. From 10 sec. to 70 sec. calls were dropped almost at the same rate, for 10 seconds time intervals, but form 70 sec. and longer we see a decrease in the pattern (shown below).Graph3

we will continue this analysis with comparing the frequency of abandoned calls vs. different sections of the call menu to understand the reaction of the calls towards the menu.


To Get a 1-2-1 Response, Call 2-1-1

United Way of Metro Atlanta offers serious commitment to their customers to maintain full satisfaction and ensure that all individuals have the opportunity to thrive and be part of a prospering community. The United Way of Metropolitan Atlanta was the first to introduce a 2-1-1 service in 1997. 2-1-1 is completely operated by private non-profit community-service organizations. The organization offers variety of informational services that ranges from debt counseling and financial assistance to emergency food and homeless services.

We are a committed team of GATECH students that will work closely with 211 United Way of Greater Atlanta to help them reach their optimal goal to provide the best services to the disadvantaged individuals of the community. Our team consists of Hamid Mohammadi, a master student in the program of Statistics; Richard Huckaby, a student in computer engineering, and myself, Fatheia Ahmeda; an Applied Mathematics student.

We set an appointment with Mr. Zubler, the director of the United Way 2-1-1 on the 20th of May where we met at the Atlanta main office.  Mr Zubler gave us a brief introduction to the operation of his office that included a tour of the place. He was very helpful and patient in dealing with all our inquiries and he provided us with samples data that included records of day to day activities. The second day, Thursday 21, Hamid went back to the office and spent all his afternoon with the director to explore more data.  As a first step in working and analyzing the data we have  constructed a visualization diagram that displays the distribution of call volume for the top 20 counties in Georgia .

We are planning several approaches to work with the available data to minimize wait time for callers and optimize the Interactive Voice Response system (IVR) based on the analysis of the data.  Automating the routine customer service interactions is the most realistic solution to solve the wait time issue. The automated voice menu will reduce the need for agents to handle the call manually and thus decrease the wait time by a considerable time.

Several critical pieces of data are disparate across different files which makes merging the data for analysis challenging. For example, the file that contains the duration for each call is separate from the one containing the origin/location of the call and there is no unique identifier to join the records from each file.
As a team we are excited to face the challenge and exert our best effort to finish our tasks.