albert::UsageScoring Class Reference

#include <albert/usagescoring.h>

Description

Modifies match scores according to user usage history and preferences.

Public Member Functions
double	modifiedMatchScore (const ItemKey &key, double match_score) const
	Returns the modified match_score for an item identified by key.
void	modifyMatchScores (const QString &extension_id, std::vector< albert::RankItem > &rank_items) const
	Modifies the match score of rank_item for an item identified by key in-place.

Public Attributes
bool	prioritize_perfect_match
	If true perfect matches should be prioritized even if their usage score is lower.
double	memory_decay
	The exponential decay applied to usage scores based on recency.
std::shared_ptr< const std::unordered_map< ItemKey, double > >	usage_scores
	The usage scores.

Member Data Documentation

memory_decay

double albert::UsageScoring::memory_decay

The exponential decay applied to usage scores based on recency.

This value adjusts the influence of recent item activations using a geometric weighting scheme: each activation contributes a weight of 1 / (memory_decay^recency). A value of 1.0 disables decay, assigning equal weight to all activations and the score of an item is the sum of its activations (Most Frequently Used). A value of 0.5 implies that for any activation a_i in history the sum of all older activations can not exceed the weight of a_i (Most Recently Used). Valid range: [0.5, 1.0]

---

The documentation for this class was generated from the following file:

• workspace/albert/include/albert/usagescoring.h