SparseEmbedder: Sparse Embedder (BM25/TF-IDF)

In VectrixDB: Lightweight Vector Database with Embedded Machine Learning Models

Sparse Embedder (BM25/TF-IDF)

Description

Generates sparse BM25 embeddings for keyword search

Public fields

vocab

Vocabulary

language

Language setting ("en" or "ml")

Methods

Public methods

• SparseEmbedder$new()

• SparseEmbedder$fit()

• SparseEmbedder$embed()

• SparseEmbedder$query_terms()

• SparseEmbedder$clone()

---

Method new()

Create a new SparseEmbedder

Usage

SparseEmbedder$new(language = "en")

Arguments

language

Language behavior ("en" = ASCII-focused, "ml" = Unicode-aware)

---

Method fit()

Fit the embedder on a corpus

Usage

SparseEmbedder$fit(texts)

Arguments

texts

Character vector of texts

---

Method embed()

Embed texts to sparse vectors

Usage

SparseEmbedder$embed(texts)

Arguments

texts

Character vector of texts

Returns

Sparse matrix of BM25 scores

---

Method query_terms()

Get term scores for a query

Usage

SparseEmbedder$query_terms(query)

Arguments

query

Query text

Returns

Named vector of term scores

---

Method clone()

The objects of this class are cloneable with this method.

Usage

SparseEmbedder$clone(deep = FALSE)

Arguments

deep

Whether to make a deep clone.

VectrixDB documentation built on Feb. 20, 2026, 5:09 p.m.