Week 1 — Data Science and AI

1. Environment Setup

Before starting the assignment, we prepared the Python environment.

2. Steps Performed

Step 1: Import Libraries

import pandas as pd
from tabulate import tabulate

Project: E-Commerce Recommendation System
Dataset: Ecommerce Customer Service Satisfaction (source: Kaggle)

📌 Dataset Information

Dataset Name Total Rows Total Columns
Customer_support_data 85,907 20

Source: Kaggle — customer support / CSAT dataset used for satisfaction analysis and recommendation experiments.

📌 Column Descriptions

Column NameDescription
Unique idUnique record identifier (UUID)
channel_nameContact channel (Inbound, Outcall, Email, etc.)
categoryHigh-level issue category
Sub-categorySub-category or more specific issue type
Customer RemarksFree-text remarks / customer comments
Order_idOrder identifier related to the issue (if applicable)
order_date_timeDate/time of order placement
Issue_reported atTimestamp when issue was reported
issue_respondedTimestamp when agent responded
Survey_response_DateDate when customer submitted the CSAT survey
Customer_CityCity of the customer (if provided)
Product_categoryProduct category associated with the ticket
Item_pricePrice of the item (if available)
connected_handling_timeCall/chat handling duration (seconds/mins)
Agent_nameName of the handling agent
SupervisorSupervisor overseeing the agent
ManagerManager in chain of command
Tenure BucketAgent tenure category (e.g., 0-30, >90)
Agent ShiftShift of the agent (Morning/Evening/Split)
CSAT ScoreCustomer satisfaction score (e.g., 1–5)

📌 Dataset Sample (First 10 Rows)

S.No. Unique idchannel_namecategorySub-category Customer RemarksOrder_idorder_date_timeIssue_reported at issue_respondedSurvey_response_DateCustomer_CityProduct_category Item_priceconnected_handling_timeAgent_nameSupervisor ManagerTenure BucketAgent ShiftCSAT Score
1 7e9ae164-6a8b-4521-a2d4-58f7c9fff13fOutcallProduct QueriesLife Insurance nanc27c9bb4-fa36-4140-9f1f-21009254ffdbnan01/08/2023 11:13 01/08/2023 11:4701-Aug-23nannannannan Richard BuchananMason GuptaJennifer NguyenOn Job TrainingMorning5
2 b07ec1b0-f376-43b6-86df-ec03da3b2e16OutcallProduct QueriesProduct Specific Information nand406b0c7-ce17-4654-b9de-f08d421254bdnan01/08/2023 12:52 01/08/2023 12:5401-Aug-23nannannannan Vicki CollinsDylan KimMichael Lee>90Morning5
3 200814dd-27c7-4149-ba2b-bd3af3092880InboundOrder RelatedInstallation/demo nanc273368d-b961-44cb-beaf-62d6fd6c00d5nan01/08/2023 20:16 01/08/2023 20:3801-Aug-23nannannannan Duane NormanJackson ParkWilliam KimOn Job TrainingEvening5
4 eb0d3e53-c1ca-42d3-8486-e42c8d622135InboundReturnsReverse Pickup Enquiry nan5aed0059-55a4-4ec6-bb54-97942092020anan01/08/2023 20:56 01/08/2023 21:1601-Aug-23nannannannan Patrick FloresOlivia WangJohn Smith>90Evening5
5 ba903143-1e54-406c-b969-46c52f92e5dfInboundCancellationNot Needed nane8bed5a9-6933-4aff-9dc6-ccefd7dcde59nan01/08/2023 10:30 01/08/2023 10:3201-Aug-23nannannannan Christopher SanchezAustin JohnsonMichael Lee0-30Morning5
6 1cfde5b9-6112-44fc-8f3b-892196137a62EmailReturnsFraudulent User nana2938961-2833-45f1-83d6-678d9555c603nan01/08/2023 15:13 01/08/2023 18:3901-Aug-23nannannannan Desiree NewtonEmma ParkJohn Smith0-30Morning5
7 11a3ffd8-1d6b-4806-b198-c60b5934c9bcOutcallProduct QueriesProduct Specific Information nanbfcb562b-9a2f-4cca-aa79-fd4e2952f901nan01/08/2023 15:31 01/08/2023 23:5201-Aug-23nannannannan Shannon HicksAiden PatelOlivia Tan>90Morning5
8 372b51a5-fa19-4a31-a4b8-a21de117d75eInboundReturnsExchange / Replacement Very good88537e0b-5ffa-43f9-bbe2-fe57a0f4e4aenan01/08/2023 16:17 01/08/2023 16:2301-Aug-23nannannannan Laura SmithEvelyn KimuraJennifer NguyenOn Job TrainingEvening5
9 6e4413db-4e16-42fc-ac92-2f402e3df03cInboundReturnsMissing Shopzilla app and it's all coustomer care services is very good service provided all time e6be9713-13c3-493c-8a91-2137cbbfa7e6nan01/08/2023 21:03 01/08/2023 21:0701-Aug-23nannannannan David SmithNathan PatelJohn Smith>90Split5
10 b0a65350-64a5-4603-8b9a-a24a4a145d08InboundShopzilla RelatedGeneral Enquiry nanc7caa804-2525-499e-b202-4c781cb68974nan01/08/2023 23:31 01/08/2023 23:3601-Aug-23nannannannan Tabitha AyalaAmelia TanakaMichael Lee31-60Evening5

Note: many columns contain nan because those fields are empty in the raw export.

📌 Missing Values per Column

Column NameMissing Values
Unique id0
channel_name0
category0
Sub-category0
Customer Remarks57,165
Order_id18,232
order_date_time68,693
Issue_reported at0
issue_responded0
Survey_response_Date0
Customer_City68,828
Product_category68,711
Item_price68,701
connected_handling_time85,665
Agent_name0
Supervisor0
Manager0
Tenure Bucket0
Agent Shift0
CSAT Score0

📌 Dataset Summary

Total RowsTotal ColumnsDuplicate RowsMissing CellsNumeric ColumnsCategorical Columns
85,907200435,995317

Milestone: dataset loaded and explored. Next step: data cleaning (handle missing values, convert types, deduplicate, and prepare for modeling).