tf idf - Calculate Cosine Similarity Using java Program -

I have a problem calculating the equality measurement for developing a search engine for my final project.

In Java, I have to use the TF IDF + cosine equality and I have no idea how to calculate it.

For your information, I have my own database that has 811 documents

Calculate the cosine equality of vector u and v, consider u and v as normal and then get the product of u and v. This means that the vectors are the same and there are numerical vectors (see) coding such operations are trivial, and some people did this for you, like here

In a search engine, cosine parity How much object can be a measure of one match. Your query is an object A, calculate the cosine parity for all the object b in your database / store / whatever, b objects sorted by decreasing equality.

If your objects are numerical vectors, then it is easy. If not, you have to prepare a way to convert your objects into digital vectors. For example, for text data, the vector may contain the number of keywords generated in the text, it is called "word model bag" (see) completely ignores such a model that the words are interchangeable with each other How to relate to a clever way, which takes simple relationships between the words of the account, computing for a given lesson may be that a given word follows another, There is a Swedish representation. The vector is then a vector of probabilities, which adheres to the word x y.