Search-Based Software Engineering for Self-Adaptive Systems: Survey, Disappointments, Suggestions and Opportunities

SE/SAS Domain Expertise	Reasons
Feature model $`\Rightarrow`$ R&F (10)
the variability of the software.
Markov model $`\Rightarrow`$ R&F (7)
of the software states.
Goal model $`\Rightarrow`$ R&F (5)
the stakeholders’ needs.
Tactics $`\Rightarrow`$ R&F (5)
using prior expertise.

$`\Rightarrow`$ R&F (2)
sources that adapt the software.
Feature model $`\Rightarrow`$ O (2)
and improve efficiency.
Tactics $`\Rightarrow`$ O (2)
space of adaptation.

$`\Rightarrow`$ O (1)
Goal model $`\Rightarrow`$ O (1)
based on requirements.
Seeding $`\Rightarrow`$ C (4)
towards expected adaptation.
Markov model $`\Rightarrow`$ S (2)

Reasons of leveraging SE/SAS domain expertise and their specializations in different parts of search algorithms on SBSE for SAS.

R, F, O, C and S denote representation, fitness function, operator, candidate solution and solution selection, respectively.

Number in the bracket indicates how many studies are involved.

ID	Item	RQ
$`I_1`$	Author(s)	N/A
$`I_2`$	Year	N/A
$`I_3`$	Title	N/A
$`I_4`$	Venue (journal or conference)	N/A
$`I_5`$	Citation count	N/A
$`I_6`$	Selected search algorithm(s) and reasons	RQ1
$`I_7`$	# algorithm(s) compared quantitatively	RQ1
$`I_8`$	SAS problem(s) to be searched	RQ1,RQ2
$`I_{9}`$	Multi-objectivity formalization and reasons	RQ2
$`I_{10}`$	Formalization assumptions	RQ2
$`I_{11}`$	Quality indicator for multiple objectives and reasons	RQ3
$`I_{12}`$	Domain information in search and reasons	RQ4
$`I_{13}`$		RQ4
$`I_{14}`$	Subject SAS(s) used and reasons	RQ5

Data collection items.