Presentation from the IWSM Mensura 2015 conference held October 5-7 in Cracow, Poland

1. EARLY OR RAPID COSMIC FSM
BY USING
APPROXIMATION APPROACHES
International Workshop on Software Measurement 2015
Frank Vogelezang
COSMIC MPC

2. Agenda
 When do you need approximation
 Count, Compute, Judge
 Scaling
 Classification
 Localization
 The different approaches

3. When do you need approximation

4. Reasons for approximation
 Speed
 Timing
 Quality

5. Rapid size measurement
 High-level size estimate
 The sooner, the better
 Board-level decisions

6. Early size measurement
 Not all details are documented yet
 Needed now
 Budget decisions

7. Approximate size measurement
 Not all documentation is available
 Quality of documentation is poor
 Portfolio decisions

8. Count, compute, judge

10. How many people
Tickets scanned
41.392
Suites & boxes
24.802
Staff
1.593
67.787 people
COUNT
Upper ring
~5 sections x
~1.500 seats
Lower ring
~16 sections x
~2.000 seats
Suites & boxes
~25.000 seats
~64.500 people
COMPUTE
Capacity: 114.500
45% full
~ 51.500 people
JUDGE
Steve McConnell - Software Estimation, Demystifying the Black Art

11. Count, Compute, Judge
COSMIC Standard Count
Current
approximation approaches Compute
New approaches Judge &
Compute

12. Scaling

13. Scaling – overview

14. Scaling – high-level

15. Scaling – zooming-in

16. Scaling – details

17. Scaling – Level of Granularity

18. Classification

19. Classification

20. Classification

21. Localization

22. Localization
 All approaches are based on documents
 Content can vary from place to place
 Local calibration is necessary
 Local means that calibration is done in an
environment that is representative

23. The localization principle
 Define high-level software artifacts
 Verify that they are representative
 Artifacts must be similar in size or
classifiable into size bands
 Determine the scaling factor
 Always give an indication of precision
 Verify the local results

24. The different approaches
 Average functional process
 Fixed size classification
 Equal size bands
 Average use case
 Early & Quick
 EASY
 Textual requirements
 EPCU

25. The case:Web Advice Module
Early specification shows two Use Cases:
1. Web Advice
2. Support processes
This module aims to help customers
seeking advice via the web on whether to
rent or obtain a mortgage to buy a house

26. UC1 – Web Advice
 On the start page the customer is presented with a short
introduction text and five propositions with which he can agree,
disagree or choose neutral. Based on the set of choices the
customer gets a general advice on the best option and specific
texts for each choice that has an answer in the category of the
general advice. The application contains business rules that define
which answer leads to which type of general advice and what
specific text is displayed that relates to a specific choice.
 In addition, the customer may require advice on renting or on a
mortgage to buy a house. To help the customer, he may send a
request for an advice session from the Web Advice Module. The
customer enters personal and contact data in a request form.
 From the Web Advice Module the customer can access an existing
mortgage assessment module.

27. UC2 – Support processes
 When the Web Advice Module detects erroneous processing, this
must be communicated to the user as an error message in a
separate window.
 If the customer has not used any functionality of the pages of the
Web Advice Module for more than five minutes, a pop-up message
will appear with a message that the customer has been inactive for
too long and that the entered data and answers are not available
anymore.
 All texts must be editable without the assistance of a programmer.

28. Average functional process
Scaling from functional process to CFP
4.880 CFP
610 Functional Processes
8 CFP/Functional Process

29. Early estimate
Average functional process approximation
 Identify the functional processes
 Multiply by 8 CFP/Functional Process

30. Fixed size classification
Assign a category to a functional process:
Small 5 CFP
Medium 10 CFP
Large 15 CFP
. . . .
 To estimate:
Classify in which band a FP belongs

31. Early estimate
Fixed size classification approximation
 Identify the functional processes
 Classify each process as:
 Small 5 CFP
 Medium 10 CFP
 Large 15 CFP

32. Equal size bands
 Count a sample of software
 Sort the functional processes on size
 Divide the total size in bands
 Calculate the size of an average
functional process in each band
 To estimate:
Classify in which band a FP belongs

33. Equal size bands – example
96 CFP
12 FP
4 bands
(4 x 3 + 2 x 6) / 6 = 4
(6 + 2 x 9) / 3 = 8
(2 x 12) / 2 = 12
24
0
3
6
9
12
15
18
21
24
B1 B2 B3 B4
S
M
L
XL

34. Early estimate
Equal size bands approximation
 Identify the functional processes
 Classify each functional process as:
 Small 4.8 CFP
 Medium 7.7 CFP
 Large 10.7 CFP
 Very large 16.4 CFP

35. Average Use Case
Scaling from Use Case to CFP
8 CFP/Functional Process
3.5 Functional Process/Use Case
28 CFP/Use Case

36. Early estimate
Average use case approximation
 Identify the use cases
 Multiply by 28 CFP/Use Case

37. Early & Quick
Two levels of classification

38. Early estimate
Early & Quick approximation
 Identify the processes
 Classify each process as:
 Functional Process
 Typical Process
 General Process
 Macro Process
 Look up the corresponding size estimate
 Small
 Medium
 Large
 Very large

39. EASY
Two types of classification
EASY is the acronym for EArly & SpeedY

40. Early estimate
EASY approximation
 Identify the functional processes
 Classify each functional process as:
 Small
 Medium
 Large
 Complex
 Look up the corresponding size estimate
 Little unknown
 Unknown (No FUR)

41. Approximation methods
Emerging or new approaches
Beyond

42. Textual requirements
 Count informally written requirements
per functional process
 Store requirements and size as reference
 Divide functional processes into sets of
fuzzy size classes
 Train a text classification algorithm to the
linguistic features of the reference set

43. EPCU
EPCU is the acronym for Estimation of Projects in a Context of Uncertainty
“The Uncertainty: it is not possible to measure it,
however it is possible to contextualize it”

44. 44
Generador de Estimados
From seemingly exact experts

45. 45
Generador de Estimados
To focused uncertainty
More details in the
15.30 workshop by

46. Early estimate
EPCU approximation
 Identify the functional processes
 Visit www.mepe.com.mx
 Estimate each functional process based on:
 Functional Process Size
 Level of Object of Interest

47. MEPE estimation models

48. The EPCU Estimation Model

49. EPCU Size classification
Functional process Classification
(linguistic
values)
Functional
process size
(value
assignment)
Start page Large Many
Advice Medium Few
Mortgage
assessment Small Average
Error handling Small Low
Inactivity Small Low
Editable texts Small Low
Translate to a
scale of 0 – 5
 Small 2
 Medium 2,5
 Large 4
 Low 0.5
 Few 2
 Average 3
 Many 3.5

50. EPCU Measurement
Start page
 Functional classification: Large
 Functional process size: Many

51. EPCU Measurement
Advice
 Functional classification: Medium
 Functional process size: Few

52. EPCU Measurement
Mortgage Assessment
 Functional classification: Small
 Functional process size: Average

53. EPCU Measurement
Error handling | Inactivity | Editable texts
 Functional classification: Small
 Functional process size: Low

54. EPCU Measurement results
Functional process Classification
(linguistic
values)
Estimated
Functional Size
using ‘Equal Size
Bands’ approach
Functional
process size
(value
assignment)
Presence of OOI
related to the
functional
process (value
assignment)
Approximation
using EPCU
Model
Start page Large 4 Many 3.5 15.09
Advice Medium 2.5 Few 2 8.53
Mortgage
assessment Small 2 Average 3 10.70
Error handling Small 2 Low 0.5 4.59
Inactivity Small 2 Low 0.5 4.59
Editable texts Small 2 Low 0.5 4.59
Total 48.09

55. Approximation results
 Average Functional Process 48 CFP
 Fixed Size Classification 45 CFP
 Equal Size Bands 43.5 CFP
 Average Use Case 56 CFP
 Early & Quick 39.6 CFP
 EASY 56.1 CFP
 EPCU 48.09 CFP

56. Approximation versus Detailed
 Detailed functional size 42 CFP
 PERT on approximation
(39.6 + 4 x 48 + 56.1)
= 47.95 CFP
6

57. Recap
 Three reasons for approximation
Speed, Timing or Quality
 Six established approaches available
 New approaches in development
 Always calibrate locally
 Guideline for Early or Rapid COSMIC FSM
cosmic-sizing.org/?p=3244

58. ANSWERS!
QUESTIONS?

59. www.cosmic-sizing.org