Recommendation System

In [1]:

#!pip install Levenshtein
#!pip install sentence_transformers

#!pip install Levenshtein #!pip install sentence_transformers

In [2]:

import intellikit as ik
import pandas as pd

import intellikit as ik import pandas as pd

/home/runner/.local/lib/python3.11/site-packages/sentence_transformers/cross_encoder/CrossEncoder.py:13: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html
  from tqdm.autonotebook import tqdm, trange

In [3]:

# Define your DataFrame on which the recommendation system is going to be based
df = pd.DataFrame({
    'feature1': ['apple', 'orange', 'banana', 'grape'],
    'feature2': ['red', 'green', 'blue', 'yellow'],
    'feature3': ['small', 'large', 'medium', 'small'],
    'feature4': [1, 2, 3, 4],
    'feature5': [10, 20, 30, 40]
})

#Define your query 
query = pd.DataFrame({
    'feature1': ['apple'],
    'feature2': ['yellow'],
    'feature3': ['small'],
    'feature4': [3],
    'feature5': [30]

})

#Define you similarity calculation methods for your project
hamming = ik.sim_hamming
levenshtein = ik.sim_levenshtein
level = ik.sim_level
abs_diff = ik.sim_difference

# Assign the appro*priate similarity calculation functions to each feature
similarity_functions = {
    'feature1': hamming,
    'feature2': levenshtein,
    'feature3': level,
    'feature4': abs_diff,
    'feature5': abs_diff
}

# Define your DataFrame on which the recommendation system is going to be based df = pd.DataFrame({ 'feature1': ['apple', 'orange', 'banana', 'grape'], 'feature2': ['red', 'green', 'blue', 'yellow'], 'feature3': ['small', 'large', 'medium', 'small'], 'feature4': [1, 2, 3, 4], 'feature5': [10, 20, 30, 40] }) #Define your query query = pd.DataFrame({ 'feature1': ['apple'], 'feature2': ['yellow'], 'feature3': ['small'], 'feature4': [3], 'feature5': [30] }) #Define you similarity calculation methods for your project hamming = ik.sim_hamming levenshtein = ik.sim_levenshtein level = ik.sim_level abs_diff = ik.sim_difference # Assign the appro*priate similarity calculation functions to each feature similarity_functions = { 'feature1': hamming, 'feature2': levenshtein, 'feature3': level, 'feature4': abs_diff, 'feature5': abs_diff }

In [4]:

# Applying the methods and weights and retrieving the top results using the linear retriever
feature_weights = {
    'feature1': 0.4,
    'feature2': 0.2,
    'feature3': 0.1,
    'feature4': 0.1,
    'feature5': 0.2
}

top_n = 2  # Number of top similar results to return
top_similar_cases_linear = ik.linearRetriever(df, query, similarity_functions, feature_weights, top_n)
print("Top similar cases (Linear):")
print(top_similar_cases_linear)

# Applying the methods and weights and retrieving the top results using the linear retriever feature_weights = { 'feature1': 0.4, 'feature2': 0.2, 'feature3': 0.1, 'feature4': 0.1, 'feature5': 0.2 } top_n = 2 # Number of top similar results to return top_similar_cases_linear = ik.linearRetriever(df, query, similarity_functions, feature_weights, top_n) print("Top similar cases (Linear):") print(top_similar_cases_linear)

Top similar cases (Linear):
  feature1 feature2 feature3  feature4  feature5
0    apple      red    small         1        10
3    grape   yellow    small         4        40