Building an Article-Recommendation-system:

• It involves building recommendation system which recomends similar articles to the readers.

Dataset:

• It involves 2 features:
  • 1st : Article - entire content of the article
  • 2nd : Title of the article

Methodology:

• using TEXT SIMILARITY to determine how the two text documents close to each other in terms of their context or meaning.
  • text similarity metrics are Cosine similarity, Euclidean distance and Jaccard Similarity,...
• The common way to compute the Cosine similarity :
  • First we need to count the word occurrence in each document.
  • To count the word occurrence in each document, we can use CountVectorizer or TfidfVectorizer functions that are provided by Scikit-Learn library.
  • cosine similarity values -- 0: less similar, 1: more similar

Steps:

• Dump all the articles in a single list:
  • data["Article"].tolist()
• Do Vectorization: Convert a collection of raw documents to a matrix of TF-IDF features:
  • from sklearn.feature_extraction import text
  • vectorization = text.TfidfVectorizer(input=articles, stop_words="english")
  • X = vectorization.fit_transform(articles)
• Applying Cosine similarity:
  • from sklearn.metrics.pairwise import cosine_similarity
  • cos_sim = cosine_similarity(X)

Creating a Recommendation column(new):

• choosing the top 4 more correlated articles.
• using argsort() : sorted + gets the corresponding Index of the values.
• data["Title"].loc[[1, 33, 21, 5, 7]] : gets the corresponding title of the Index.

Conclusion:

• To create an articles recommendation system, we need to focus on "content" rather than user interest.
• Cosine similarity is a method of building recommendation systems based on the content.
• It is used to "find similarities" between two different pieces of "text" documents.