mhdzaid/text-search

Text Search

How To Run

• This is a Spring Boot Maven project using java 8.
• Project is running on port 8080.
• The API call is endpoint is

    http://localhost:8080/api/v1/words?similar=

• The word whose frequency and similar words are to be determined is the Request Parameter similar= e.g similar=Word
• The Request Body should be

    {
        "text":"Word Word Word word"
    }

• The Response should be

    {
        "similarWords": [
            "word"
        ],
        "wordFrequency": 3
    }

Scope of Project

• I've implemented text search using Apache Lucene which has a lot of features.
• The time taken for me to develop this API was 3 hours which mostly involved reading Apache Lucene docs and testing different implementations.
• It is intentionally case-senstive e.g word and Word are currently considered as similar but not equal.
• The reason for choosing Apache Lucene is that even if we want to make changes later on e.g to add case sensitivity, it would be easy to do so by just changing the tokenizer.

How to bring solution further

• I took the liberty to make an extra change of disregarding commas while detecting words.
• Currently the API can detect words regardless of commas, e.g in text word, Word frequency of word would be 1;
• Given more time I would have liked to test it using further punctuations like . ? : " " etc.