Fostering compliance through automation

1
Directorate of Income Tax (Investigations) Delhi
Fostering compliance through automation
21-04-2007
2
Presentation outline

• Background
• Automation in Indian direct tax System
• Main databases and their elements
• Data mining
• Tools for non-intrusive anti-tax evasion work

3
Background

• World class tax systems rest on
• Moderate tax rates
• Simple and transparent procedures
• Quality taxpayer services for enabling voluntary
  compliance
• Effective deterrence
• Requires full automation, centralisation and
  functional specialisation

4

• Automation in Indian tax system

5
Automation of Indian tax system the big picture

• Objectives
• Taxpayer services on 24/7 basis
• Compliance management through non-discretionary
  automated tools
• Effective deterrence through intelligent
  databases
• Single national database connecting all offices
  through dedicated National network
• Contd.-

6
--- the big picture (contd.)

• Process reform for -
• Jurisdiction-free filing,
• Centralised mass processing of returns at a few
  locations across the country
• Refund cheques through Refund banker
• Computer assisted automated selection of cases
  for scrutiny
• Data mining using departmental/other databases
  for identifying patterns and exceptions

7

• Main departmental databases

8
Main departmental databases and their elements

• 1. Taxpayer identification- PAN
• Name, fathers name, address, nature of
  business, partner/ director
• 2. Tax payment/ refund data- OLTAS
• PAN, name, Major/minor head, date, amount
  (historical data)
• 3. Data from returns of income- AST
• Head wise incomes, deductions, rebate, pre-paid
  taxes, MAT (historical data)

9
Main departmental databases and their elements (2)

• 4. Data of tax deduction at source-TDS
• TAN/ Deductor, PAN/ deductee, amount paid, date,
  nature of payment, TDS
• Assessee-wise ledger of TDS/TCS (F.26AS)
• 5. Data of specified high value transaction - AIR
  
• PAN/Name, amount, date, nature, counter party
• Assessee-wise Transaction statement (ITS)
• 6. Others- STT, CIB, BCTT

10
External databases

• Registered companies- MCA 21
• Information of compliances/ filings under
  Companies Act
• Mobile phone Vs PAN
• Off-market securities transactions
• Research databases- CMIE, Capital line
• Future databases- TIN xys, State VAT, FIU

11

• Data Warehousing and data mining

12
Need for data mining

• Large volumes of electronic information of
  different types, in different formats, flowing
  into the department
• Manual analysis, co-relation of large diverse
  databases not possible
• Need to identify patterns, locate exceptions,
  convert into actionable intelligence

13
Data Warehousing and data mining techniques

• Data warehouse is a subject-oriented, integrated,
  historical, and non-volatile collection of data
  for use in decision making.
• Data warehousing is a technique for assembling
  data from various sources for getting a single,
  detailed view of part or all of a business/
  organisation
• Data mining is the process of extracting valid,
  previously unknown, and ultimately comprehensible
  information and linkages from large databases for
  critical business decisions

14

• Profile builder
• A non-intrusive tool for anti-tax evasion work

15
Main features of Profile builder

• A data mining tool that enables matching of large
  databases of 5 crore records, in different
  formats,
• can match transaction data in AIR, CIB, STT with
  identity data in PAN and income data in AST, and
  tax data OLTAS, e-TDS
• Matches names and addresses on phonetic as well
  as pattern recognition methods, using a
  customised search engine
• Search can start simply from a name or PAN or
• phone number

16
Main features of Profile builder (2)

• Ranks search results according to the probability
  of matching
• Creates 360 degree profile of a taxpayer putting
  together listing his-
• Group concerns, associates
• Family tree
• Transactions in AIR/CIB/STT etc
• Disclosed incomes as per returns and
• Taxes paid /deducted as per OLTAS / TDS

17
Schematic representation of Profile builder
Director in companies?
Partner in firms?
PAN
Identity particulars?
Group concerns/ associates
Is his father assessed to tax?
PAN
Profile
AST
Income details
Are his brothers assessed to tax?
PAN
OLTAS
Investments / Expenditure
Prepaid taxes
Are his children assessed to tax?
PAN
e-TDS
CIB data
STT
AIR data
18
Architecture of Profile builder
Data sources
Investigation Wing
PAN
Combined Data
Combined Data with PAN
Investigation Wing
Combined Data without PAN
Valuable Information
Data cleaning
Data Mining
19

• Thanks

20
Difficulties in name matching

• 1. Variations in writing the same name
• Spelling G Venu
  Gopal Rao G Veni Gopala Rau
• Abbreviation G.V.G.Rao
  G.V.Gopal Rao
• Prefix Mr.
  G.Venugopal Rao Sri. Sree,
• Punctuation G.Venu
  Gopal , Rao
• Order of writing G Venu Gopal
  Rao ,Venu Gopal Rao G
• 2. Missing/ extra parts G Venu Gopal, G
  venu Rao, G.Gopal rao
• 3. Split name/concatenation G Venugopal Rao ,G
  Venugopalrao
• 4. Combination of all above
• Source Mr. Gopal Rau G
• PAN Gouravelly Venu Gopal Rao