What is Term Frequency?
Term Frequency (TF) counts how many times a term appears in a document. In ranking, higher frequency usually signals higher importance—but it’s normalized to avoid bias toward long documents.
How It Works (quick)
- Count terms: Raw term counts per document.
- Normalize: Divide by document length to avoid bias.
- Weight: Used in formulas like TF-IDF or BM25.
- Impact: More occurrences → higher relevance, up to a saturation point.
Why It Matters in E-commerce
- Ranking: Queries like “leather boots” rank documents with higher occurrences of those words.
- SEO: Helps understand keyword density in product descriptions.
Best Practices
- Avoid keyword stuffing; focus on natural usage.
- Use unique attributes to boost relevance (brand, model).
- Pair with semantic methods for meaning beyond raw counts.
Summary
Term frequency is a basic relevance signal. It works best when combined with normalization and other ranking factors.
