software-engineering-research

Imported from wentorai/research-plugins (skills/domains/cs/software-engineering-research/SKILL.md). Install upstream with npx skills add wentorai/research-plugins --skill software-engineering-research. Copyright stays with the author.

Software Engineering Research Guide

Navigate the landscape of software engineering research, including key subfields, methodologies, datasets, benchmarks, and top venues.

SE Research Subfields

Research Methodologies in SE

Controlled Experiments

Testing a specific hypothesis with treatment and control groups:

Example: Does AI code completion improve developer productivity?

Design:
- Participants: 60 professional developers
- Treatment: IDE with AI code completion enabled
- Control: IDE with AI code completion disabled
- Task: Complete 5 programming tasks of varying difficulty
- Metrics: Task completion time, code correctness, lines of code
- Analysis: Mixed-effects linear model with participant as random effect

Threats to validity:
- Internal: Learning effect (counterbalance task order)
- External: Lab setting may not reflect real development
- Construct: "Productivity" operationalized as speed + correctness

Mining Software Repositories (MSR)

Analyzing data from version control, issue trackers, code review systems:

# Example: Analyze commit patterns using PyDriller
from pydriller import Repository

repo_url = "https://github.com/apache/kafka"

commit_data = []
for commit in Repository(repo_url, since=datetime(2023, 1, 1),
                          to=datetime(2023, 12, 31)).traverse_commits():
    commit_data.append({
        "hash": commit.hash[:8],
        "author": commit.author.name,
        "date": commit.committer_date,
        "files_changed": commit.files,
        "insertions": commit.insertions,
        "deletions": commit.deletions,
        "message": commit.msg[:100]
    })

df = pd.DataFrame(commit_data)
print(f"Total commits in 2023: {len(df)}")
print(f"Unique contributors: {df['author'].nunique()}")
print(f"Avg files per commit: {df['files_changed'].mean():.1f}")

Case Studies

In-depth investigation of a phenomenon in its real-world context:

Case Study Protocol (based on Yin, 2018):
1. Research questions: How do teams adopt microservices?
2. Unit of analysis: Development teams at 3 companies
3. Data sources:
   - Semi-structured interviews (8-12 per company)
   - Architecture documentation review
   - Commit history and deployment logs
   - Meeting observations
4. Analysis: Thematic analysis with cross-case comparison
5. Validity: Triangulation across data sources, member checking

Key Datasets and Benchmarks

Code Understanding and Generation

Software Engineering Process

Static Analysis Tools for Research

# Example: Using tree-sitter for AST-level code analysis
from tree_sitter import Language, Parser
import tree_sitter_python as tspython

PYTHON_LANGUAGE = Language(tspython.language())
parser = Parser(PYTHON_LANGUAGE)

source_code = b"""
def fibonacci(n):
    if n <= 1:
        return n
    return fibonacci(n-1) + fibonacci(n-2)
"""

tree = parser.parse(source_code)
root = tree.root_node

def count_nodes(node, node_type):
    """Count AST nodes of a given type."""
    count = 1 if node.type == node_type else 0
    for child in node.children:
        count += count_nodes(child, node_type)
    return count

print(f"Function definitions: {count_nodes(root, 'function_definition')}")
print(f"If statements: {count_nodes(root, 'if_statement')}")
print(f"Return statements: {count_nodes(root, 'return_statement')}")
print(f"Function calls: {count_nodes(root, 'call')}")

Code Metrics

# Common software metrics
metrics = {
    "Lines of Code (LOC)": "Total lines (including blanks and comments)",
    "Cyclomatic Complexity": "Number of independent paths (McCabe, 1976)",
    "Halstead Volume": "Based on operators and operands count",
    "Maintainability Index": "Composite of LOC, CC, and Halstead",
    "Coupling Between Objects": "Number of other classes referenced",
    "Depth of Inheritance": "Levels in class hierarchy",
    "Code Churn": "Lines added + modified + deleted per period",
    "Comment Density": "Ratio of comment lines to total lines"
}

# Calculate cyclomatic complexity using radon
# pip install radon
import subprocess
result = subprocess.run(
    ["radon", "cc", "my_module.py", "-s", "-j"],
    capture_output=True, text=True
)
print(result.stdout)

Top Venues and Impact

Tier-1 SE Venues

Systems and Security Venues

Research Tools Ecosystem