NCA-GENM Question 128
Select 2You are tasked with building a multimodal AI application that integrates text and image data for a recommendation system. For the text analysis component, you decide to use Python natural language processing tools to extract key entities from user reviews and store them efficiently for semantic search. Which combination of tools would be most suitable for this task?
- A
spaCy for named entity recognition and NumPy for efficient numerical computations
- B
spaCy for named entity recognition and a vector database like FAISS for semantic search
- C
NumPy for named entity recognition and a traditional SQL database for semantic search
- D
A vector database like FAISS for named entity extraction and spaCy for semantic search
- E
A vector database like FAISS for semantic search and spaCy for named entity recognition
Show answer and explanation
Correct answers: B, E
Explanation
To build the text analysis component of the multimodal AI application, spaCy should be used for named entity recognition as it is a powerful NLP library for such tasks. For semantic search, a vector database like FAISS is ideal because it enables fast and efficient similarity searches on high-dimensional embeddings. Combining these tools ensures both accurate entity extraction and efficient search capabilities, fulfilling the application's requirements.
- A. Incorrect.
spaCy is excellent for named entity recognition, but NumPy is generally used for numerical computations rather than semantic search tasks, making this combination incomplete for the given scenario.
- B. Correct.
This is a correct option as spaCy is well-suited for named entity recognition, and FAISS is designed for efficient semantic search using vector similarity.
- C. Incorrect.
NumPy is not designed for named entity recognition or semantic search, and a traditional SQL database is not optimized for vector-based semantic search.
- D. Incorrect.
FAISS is not used for named entity extraction, and spaCy is not a tool for semantic search, so this combination is incorrect.
- E. Correct.
This is a correct option because FAISS can handle semantic search tasks effectively using vector embeddings, while spaCy is ideal for named entity recognition.