Cosine Similarity Algorithm | Machine Learning

[Arko-Sengupta]

Cosine Similarity Algorithm

Overview

Introduces a New Implementation of the Cosine Similarity Algorithm in the Cosine_Similarity class. Cosine Similarity is a widely used metric in Natural Language Processing and Information retrieval to measure the similarity between two texts based on their Vector Representations.

Key Features

• Vector Representation: Utilizes SpaCy's pre-trained Word Embeddings to convert text into Vectors.
• Tokenization: Breaks down input text into lowercased tokens, excluding punctuation.
• Vectorization: Converts tokens into their corresponding vectors using SpaCy's embeddings.
• Mean Vector Calculation: Computes the Mean Vector for a set of Word Vectors to represent the overall text.
• Cosine Similarity Calculation: Measures the cosine of the angle between two vectors, providing a Similarity Score ranging from -1 to 1.
• Cosine Similarity Percentage: Outputs the similarity score as a percentage, facilitating easier interpretation.

Mathematical Foundation

• Dot Product: Measures the Degree of Alignment between two Vectors.

• Magnitude (Norm): Computes the length of a Vector.

• Cosine Similarity Formula:

   Cosine Similarity = (Dot Product) / (Magnitude_1 * Magnitude_2)
  

  where the result is normalized to lie between -1 and 1, with 1 indicating identical vectors, 0 indicating orthogonal vectors, and -1 indicating completely dissimilar vectors.

Usage

The Cosine_Similarity class provides methods to Tokenize, Vectorize, and calculate the Cosine Similarity between two pieces of text. It includes:

• Tokenize(text): Tokenizes the input text into lowercase tokens.
• Vectorize(tokens): Converts tokens into vector representations.
• Mean_Vector(vectors): Computes the average vector of a list of vectors.
• Dot_Product(vector1, vector2): Calculates the dot product of two vectors.
• Magnitude(vector): Computes the magnitude of a vector.
• Cosine_Similarity(vector1, vector2): Computes the cosine similarity between two vectors.
• Cosine_Similarity_Percentage(text1, text2): Calculates the similarity percentage between two texts.

Error Handling

Robust Error Handling is implemented for all operations to ensure reliability. Any issues encountered during tokenization, vectorization, or similarity calculations are logged and raised appropriately.

Benefits

• Provides an effective method for comparing textual content.
• Leverages pre-trained embeddings for accurate and efficient similarity measurement.
• Can be used in various applications including document similarity, search relevance, and recommendation systems.

[algorithms-keeper]

Click here to look at the relevant links ⬇️

🔗 Relevant Links

Repository:

• Contributing guidelines
• Project Euler solution guidelines

Python:

• Formatted string literals (f-strings)
• Type hints
• doctest
• unittest
• pytest

Automated review generated by algorithms-keeper. If there's any problem regarding this review, please open an issue about it.

---

algorithms-keeper commands and options

algorithms-keeper actions can be triggered by commenting on this PR:

• @algorithms-keeper review to trigger the checks for only added pull request files
• @algorithms-keeper review-all to trigger the checks for all the pull request files, including the modified files. As we cannot post review comments on lines not part of the diff, this command will post all the messages in one comment.

NOTE: Commands are in beta and so this feature is restricted only to a member or owner of the organization.

[QuantumNovice]

The algorithm in question is already in the repo at

Python/machine_learning/similarity_search.py

Line 143 in 729c1f9

def cosine_similarity(input_a: np.ndarray, input_b: np.ndarray) -> float:

Please rename to document similarity and double check to see if it doesn't already exists in the repository.

[Arko-Sengupta]

@QuantumNovice Resolved. Thanks for Raising!