This project is a sentence similarity checker built using natural language processing techniques and the Sentence Transformers library. It takes a user-defined sentence and compares it to up to three other sentences, providing a similarity score for each comparison. The output is displayed using a Streamlit web interface. This project can be used for various applications, such as plagiarism detection or content similarity analysis.

The problem solved by this project is the need for an automated tool to measure the similarity between two sentences. This tool can be used by businesses, researchers, or anyone who needs to compare large amounts of text quickly and accurately. The solution is a Streamlit web app that uses a pre-trained sentence-transformer model to compute cosine similarity scores between sentences

Goals/Requirements:

Develop a user-friendly interface for the web app that allows users to input sentences and view the similarity scores.

Use a pre-trained sentence-transformer model to compute cosine similarity scores quickly and accurately.

Ensure that the web app can handle large amounts of text and multiple comparison sentences.

Sure, below is an article I have written that explains how I built the tool. Give it a read to learn more about the process and the techniques used.

I built a tool using Python, Streamlit, and Sentence Transformers that computes the similarity score between a source sentence and up to three other sentences. The tool is useful for tasks like plagiarism detection and document similarity analysis. Through this project, I was able to successfully develop a functional tool that met the goals of accuracy, efficiency, and user-friendliness. The main takeaways from this project include the importance of leveraging pre-trained models for natural language processing tasks and the value of building tools that address specific needs in the domain of text analysis.