Table 1.

Component’s Features.

Category/Class	Feature	Description	Query Dependency
Similarity	Textual Similarity	Candidate code example and query likeness is measured using cosine similarity.	Yes
Popularity	Frequency	Code example method frequent occurrence of a specific in the corpus	No
	Probability	The frequency at which a method calls a code fragment. method
Code Metrics	Line length	The candidate code length in terms of lines.	No
	Number of identifiers	The average amount of identifiers associated with each LOC of a candidate fragment
	Length of call sequence	Number of times a candidate code component method is called
	Code Comment ratio	The number of comments associated with candidates LOC
	Page Ranking	A metric to compute significance of code snippets
	Fan-in	The number of times a specific code snippet calls other unique code snippets
	Fan-out	The number of times a specific code snippet is called by other code fragments
	Cyclomatic complexity	Occurrences of decisions in a code component (for, while, etc.)
Context	Context similarity	The Jaccard similarity measure is computed when a query is raised; it calculates the likeness between the method’s body and code example signature	Yes