Skip to content

User Profiling

UserProfilingRecommender

A context-aware recommender system that builds and maintains user profiles for personalized recommendations.

This recommender system creates and maintains detailed user profiles that capture user preferences, behavior patterns, and context-dependent interactions. It uses these profiles to generate personalized recommendations that adapt to different contexts.

Attributes:

Name Type Description
user_profiles Dict[int, Dict]

Detailed profiles for each user containing: - Preference vectors - Historical interactions - Context-dependent behavior patterns - Long-term and short-term interests
context_preferences Dict[int, Dict]

User preferences mapped to different contexts
preference_evolution Dict[int, List]

Temporal evolution of user preferences

Methods:

Name Description
update_profile

Updates user profile with new interaction data
analyze_context_preferences

Analyzes user preferences in different contexts
detect_preference_shifts

Identifies changes in user preferences over time
recommend

Generates personalized recommendations based on profile and context
Example

profiler = UserProfilingRecommender() profiler.update_profile(user_id=123, interaction_data={...}, context={...}) recommendations = profiler.recommend(user_id=123, context={"location": "work"})

Note
  • Profiles are automatically updated with each user interaction
  • Supports both explicit (ratings) and implicit (behavior) feedback
  • Implements drift detection for evolving user preferences
Source code in engines/contentFilterEngine/context_personalization/user_profiling.py 
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
class UserProfilingRecommender:
    """
    A context-aware recommender system that builds and maintains user profiles for personalized recommendations.

    This recommender system creates and maintains detailed user profiles that capture user preferences,
    behavior patterns, and context-dependent interactions. It uses these profiles to generate
    personalized recommendations that adapt to different contexts.

    Attributes:
        user_profiles (Dict[int, Dict]): Detailed profiles for each user containing:
            - Preference vectors
            - Historical interactions
            - Context-dependent behavior patterns
            - Long-term and short-term interests
        context_preferences (Dict[int, Dict]): User preferences mapped to different contexts
        preference_evolution (Dict[int, List]): Temporal evolution of user preferences

    Methods:
        update_profile: Updates user profile with new interaction data
        analyze_context_preferences: Analyzes user preferences in different contexts
        detect_preference_shifts: Identifies changes in user preferences over time
        recommend: Generates personalized recommendations based on profile and context

    Example:
        >>> profiler = UserProfilingRecommender()
        >>> profiler.update_profile(user_id=123, interaction_data={...}, context={...})
        >>> recommendations = profiler.recommend(user_id=123, context={"location": "work"})

    Note:
        - Profiles are automatically updated with each user interaction
        - Supports both explicit (ratings) and implicit (behavior) feedback
        - Implements drift detection for evolving user preferences
    """
    def __init__(self, user_attributes: Optional[pd.DataFrame] = None):
        """
        Initialize the user profiling recommender with optional user attributes.

        Parameters:
        - user_attributes (pd.DataFrame, optional): DataFrame containing user information.
        """
        self.user_profiles: Dict[int, Dict[str, Any]] = {}
        self.user_attributes = user_attributes

    def fit(self, user_interactions: Dict[int, List[int]]):
        """
        Build user profiles based on interactions and user attributes.

        Parameters:
        - user_interactions (dict): Dictionary mapping user IDs to lists of interacted item IDs.
        """
        for user_id, items in user_interactions.items():
            profile = {}

            # Incorporate user attributes if available
            if self.user_attributes is not None:
                user_info = self.user_attributes[self.user_attributes['user_id'] == user_id]
                if not user_info.empty:
                    user_info = user_info.iloc[0]
                    for col in self.user_attributes.columns:
                        if col != 'user_id':
                            profile[col] = user_info[col]

            # Add interacted items
            profile['interacted_items'] = set(items)
            self.user_profiles[user_id] = profile

    def recommend(self, user_id: int, all_items: set, top_n: int = 10) -> List[int]:
        """
        Generate top-N item recommendations for a given user based on their profile.

        Parameters:
        - user_id (int): The ID of the user.
        - all_items (set): Set of all available item IDs.
        - top_n (int): The number of recommendations to generate.

        Returns:
        - List[int]: List of recommended item IDs.
        """
        if user_id not in self.user_profiles:
            return []

        interacted_items = self.user_profiles[user_id].get('interacted_items', set())
        recommendations = list(all_items - interacted_items)
        return recommendations[:top_n]

__init__(user_attributes=None)

Initialize the user profiling recommender with optional user attributes.

Parameters: - user_attributes (pd.DataFrame, optional): DataFrame containing user information.

Source code in engines/contentFilterEngine/context_personalization/user_profiling.py 
38
39
40
41
42
43
44
45
46
def __init__(self, user_attributes: Optional[pd.DataFrame] = None):
    """
    Initialize the user profiling recommender with optional user attributes.

    Parameters:
    - user_attributes (pd.DataFrame, optional): DataFrame containing user information.
    """
    self.user_profiles: Dict[int, Dict[str, Any]] = {}
    self.user_attributes = user_attributes

fit(user_interactions)

Build user profiles based on interactions and user attributes.

Parameters: - user_interactions (dict): Dictionary mapping user IDs to lists of interacted item IDs.

Source code in engines/contentFilterEngine/context_personalization/user_profiling.py 
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
def fit(self, user_interactions: Dict[int, List[int]]):
    """
    Build user profiles based on interactions and user attributes.

    Parameters:
    - user_interactions (dict): Dictionary mapping user IDs to lists of interacted item IDs.
    """
    for user_id, items in user_interactions.items():
        profile = {}

        # Incorporate user attributes if available
        if self.user_attributes is not None:
            user_info = self.user_attributes[self.user_attributes['user_id'] == user_id]
            if not user_info.empty:
                user_info = user_info.iloc[0]
                for col in self.user_attributes.columns:
                    if col != 'user_id':
                        profile[col] = user_info[col]

        # Add interacted items
        profile['interacted_items'] = set(items)
        self.user_profiles[user_id] = profile

recommend(user_id, all_items, top_n=10)

Generate top-N item recommendations for a given user based on their profile.

Parameters: - user_id (int): The ID of the user. - all_items (set): Set of all available item IDs. - top_n (int): The number of recommendations to generate.

Returns: - List[int]: List of recommended item IDs.

Source code in engines/contentFilterEngine/context_personalization/user_profiling.py 
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
def recommend(self, user_id: int, all_items: set, top_n: int = 10) -> List[int]:
    """
    Generate top-N item recommendations for a given user based on their profile.

    Parameters:
    - user_id (int): The ID of the user.
    - all_items (set): Set of all available item IDs.
    - top_n (int): The number of recommendations to generate.

    Returns:
    - List[int]: List of recommended item IDs.
    """
    if user_id not in self.user_profiles:
        return []

    interacted_items = self.user_profiles[user_id].get('interacted_items', set())
    recommendations = list(all_items - interacted_items)
    return recommendations[:top_n]