Extended Boolean model¶

An information-retrieval model that relaxes exact Boolean matching by incorporating partial matching and term weights.

Version: v3 · 2026-09-19 · History
Domain-specific #: 1813
Domain group: Applied Sciences & Engineering
Origin domain: Computer Science & Software Engineering
Subdomains: Information Retrieval, Graded Boolean Retrieval → Computer Science & Software Engineering
Aliases: Extended Boolean retrieval model, P-norm model

Core Idea¶

Extended Boolean model is an information-retrieval model that relaxes exact Boolean matching by incorporating partial matching and term weights.^[1]

The extended Boolean model preserves an AND/OR query tree while replacing exact binary satisfaction with graded term weights and p-norm aggregation. Documents receive similarity scores and can be ranked; the p parameter controls how strictly each connective approximates classical Boolean logic.

Its operative boundary is not supplied by the name alone. Preserve this identity: An information-retrieval model that relaxes exact Boolean matching by incorporating partial matching and term weights. Validity boundary: Retrieval scores must combine Boolean query structure with graded term weighting under the extended model; pure vector-space ranking or exact Boolean matching is different. The entry therefore captures a reusable specialist role structure rather than a topic label, a single historical instance, or a loose analogy.

How would you explain it like I'm…

Partial-Credit Book Search

You ask a librarian for books about dogs AND parks. A strict librarian only hands you books with both. A kinder librarian gives every book a score, putting books with lots of dogs and parks first and books with only some near the end. The Extended Boolean model is that kinder librarian.

AND/OR with Partial Credit

Search systems can use words like AND and OR to describe what you want. Plain Boolean search is all-or-nothing: a document either matches exactly or it's thrown out. The Extended Boolean model keeps the AND and OR structure but gives each word a weight showing how important it is in each document. It then combines those weights into a score, so documents can be ranked from best match to worst. A setting called p decides how strict the AND and OR act.

Weighted, Ranked Boolean Retrieval

The Extended Boolean model is an information-retrieval method that softens exact Boolean matching. Your query stays a tree of AND and OR operations, but instead of each term being simply present or absent, each term in a document has a graded weight. The model combines those weights with a formula called a p-norm to give a similarity score, so results can be ranked. The parameter p tunes how closely AND and OR behave like strict logic. It differs from pure vector-space ranking, which ignores the AND/OR structure, and from exact Boolean matching, which gives no partial credit.

The Extended Boolean model relaxes exact Boolean retrieval by adding partial matching and term weights. A query remains an AND/OR tree, but documents no longer either satisfy or fail each clause: each term carries a graded weight, and connectives aggregate these weights through p-norm formulas. The result is a real-valued similarity score per document, enabling ranked output. The parameter p tunes strictness, controlling how closely each connective approximates classical Boolean AND or OR. The model sits between two neighbors: exact Boolean matching, which gives binary results, and pure vector-space ranking, which drops the query's logical structure. What makes it this model is the combination of Boolean query structure with graded weighting.

Structural Signature¶

Sig role-phrases:

the document representation — weighted term coordinates for each indexed document
the weighted query terms — term importance values in the request
the Boolean structure — the AND/OR expression tree retained from the query
the p-norm operator — the graded aggregation rule for each connective
the strictness parameter — p controlling movement from soft compensation toward Boolean behavior
the similarity score — a graded document–query match value
the ranked output — documents ordered by score rather than only accepted or rejected
the normalization — scaling that keeps term weights and scores comparable

Recognition test. A case qualifies only when the analyst can map the declared the document representation, the weighted query terms, the Boolean structure, the p-norm operator, the strictness parameter and preserve the specialist validity conditions. Shared vocabulary, a similar output, or a generic instance of one parent relation is insufficient.

What It Is Not¶

Not exact Boolean retrieval. Exact retrieval produces a binary set without partial-match ranking.
Not the vector-space model. The Boolean expression tree remains semantically active.
Not fuzzy logic in general. This is a specific p-norm information-retrieval construction.
Not probabilistic retrieval. Scores are not necessarily relevance probabilities.
Not arbitrary weighted keyword search. The model specifies how weights compose under AND and OR.

Scope of Application¶

The abstraction recurs literally within text retrieval systems that need Boolean query structure together with partial matching and ranked results. The following habitats preserve the same recognition machinery; they are not invitations to extend the name metaphorically.

Soft conjunction. documents missing one term can receive penalized nonzero scores.
Soft disjunction. multiple matched terms increase score without exact set union only.
Weighted queries. term importance influences branch aggregation.
Ranked Boolean interfaces. users retain explicit connectives but receive an ordering.
Parameter studies. p tunes strictness against recall and ranking quality.

Clarity¶

Specify term-weight normalization, the exact AND and OR formulas, p values, treatment of nested operators, and score direction. 'Boolean plus ranking' is too broad; different fuzzy and probabilistic models yield different compensation and semantics.

A practical identification audit begins with the typed roles rather than the title: establish the document representation, verify the weighted query terms, then test the remaining conditions and exclusions. If the case retains only the portable skeleton described below, it should be named through a parent abstraction rather than as Extended Boolean model.

Manages Complexity¶

The model interpolates between interpretable query logic and graded retrieval. One expression can express structural requirements while avoiding the brittle all-or-nothing boundary of classical Boolean matching.

The compression remains accountable because each simplification has a named failure condition. Disagreement can be localized to a missing role, an invalid assumption, an ambiguous measurement, or a neighboring abstraction instead of being hidden inside an unanalyzed label.

Abstract Reasoning¶

R1. Represent documents and query terms on a compatible normalized weight scale. R2. Parse the query into an explicit Boolean tree. R3. Evaluate leaves and combine them with the declared p-norm AND/OR formulas. R4. Propagate scores through nesting and rank documents. R5. Validate p and weighting choices against relevance judgments and Boolean boundary cases.

These moves separate definition, derivation, measurement, and interpretation. A formal consequence does not by itself prove that an observed case instantiates the abstraction, while an observed resemblance does not relax the formal or institutional recognition conditions.

Knowledge Transfer¶

The model transfers literally to weighted document retrieval with p-norm Boolean composition. Compositionality and similarity measure are parents; general fuzzy rules or hybrid filters are not this model without its formulas.

The transfer boundary is explicit: DOMAIN-SPECIFIC PASS / PRIME FAIL: The model applies across weighted queries and document collections where strict Boolean result sets are too coarse. Literal recognition retains the specialist vocabulary and validity conditions of information retrieval; outside that setting only broader parent operations transfer. The safe move beyond the home habitat is to carry the applicable parent relation and leave the specialist name behind unless every defining role remains literal.

Examples¶

Canonical: a soft AND query¶

For 'renewable AND storage,' a document strong on both terms ranks highest, while one weak on storage is penalized rather than categorically excluded. Increasing p makes the conjunction behave more like the minimum-like strict Boolean requirement.^[1]

Mapped back: the document representation; the weighted query terms; the Boolean structure; the p-norm operator; the strictness parameter; the similarity score.

Applied / In Practice: nested weighted retrieval¶

A query combines a high-weight phrase branch with an OR branch of synonyms. Each subtree is scored under its declared extended operator, then the root score ranks documents while retaining which query structure produced the result.^[2]

Mapped back: the weighted query terms; the Boolean structure; the p-norm operator; the ranked output; the normalization.

Structural Tensions¶

T1: Boolean interpretability vs partial compensation. Soft matching improves recall but can return a document violating an expected hard requirement. Diagnostic: Which operators are truly mandatory?

T2: p flexibility vs tuning burden. Different strictness values reshape rankings. Diagnostic: How is p validated and exposed?

T3: Term weights vs connective semantics. A large weight can dominate a branch in ways users do not expect. Diagnostic: Are score explanations available?

T4: Nested structure vs normalization. Subtree scale differences can distort root aggregation. Diagnostic: Are branches comparable?

T5: Similarity score vs relevance probability. A normalized geometric score may look probabilistic without calibration. Diagnostic: How is the output interpreted?

T6: Domain autonomy vs prime reduction. Compositionality and Similarity Measure omit the specialist objects, constraints, and validity tests named above. Diagnostic: Would retaining only the portable parent pattern still satisfy the recognition test?

Structural–Framed Character¶

The five-criterion aggregate is 0.15 (structural). The judgment is criterion-specific:

Vocabulary travels — low (0.25). The complete vocabulary remains tied to the typed roles in the Structural Signature.
Evaluative weight — low (0.00). Application carries the stated degree of normative or interpretive judgment beyond structural recognition.
Institutional origin — low (0.25). The abstraction depends to this degree on a scholarly, technical, legal, or social convention.
Human-practice bound — low (0.00). Recognition depends to this degree on organized practice, language, measurement, or institutional action.
Import versus recognize — low (0.25). Beyond its home habitat, use of the full name increasingly becomes analogy rather than literal recognition.

The portable skeleton is logical composition is relaxed into graded similarity while retaining the query's explicit operator structure. The named abstraction remains structural because that skeleton alone does not supply its specialist objects, constraints, or tests.

Structural Core vs. Domain Accent¶

Structural core: Logical composition is relaxed into graded similarity while retaining the query's explicit operator structure.

Domain accent: Document term weights, boolean query trees, p-norm and and or, strictness parameters, partial matching, normalized scores, and ranked retrieval.

Why it does not clear the prime bar: Compositionality and similarity travel; the extended Boolean model is their specific p-norm retrieval formulation. Generalization therefore routes through parent abstractions; preserving the specialist name requires the full accent.

This entry is a kind of Compositionality and presupposes Similarity Measure.

Compositionality (Compositionality). A query score is built recursively from term and connective meanings.
Similarity Measure (Similarity Measure). Documents receive graded match scores used for ranking.

Relationships to Other Abstractions¶

Current abstraction Extended Boolean model Domain-specific

Parents (2) — outgoing typed relations

Extended Boolean model is a kind of Compositionality Prime

Compositionality (prime:compositionality).
Extended Boolean model presupposes Similarity Measure Prime

Similarity Measure (prime:similarity_measure).

Hierarchy paths (3) — routes to 3 parentless roots

Extended Boolean model → is a kind of Compositionality

Show alternative paths (2)

Neighborhood in Abstraction Space¶

Extended Boolean model sits in a sparse region of the domain-specific corpus (69^th percentile for distinctiveness): few abstractions share its structure, so a faithful description tends to retrieve it precisely.

Family — Codes, Matrices & Combinatorial Problems (33 abstractions)

Nearest neighbors

Computed from structural-signature embeddings · 2026-10-25

Not to Be Confused With¶

Boolean retrieval. exact set operations with binary matching. Tell: Can partial matches receive scores?
Vector-space model. cosine or related similarity without required Boolean structure. Tell: Does an AND/OR tree govern composition?
Fuzzy retrieval. a broader family using fuzzy sets or logics. Tell: Are Salton–Fox–Wu p-norm operators used?
Probabilistic retrieval. ranking by estimated relevance probabilities. Tell: Is the score geometric or probabilistic?
BM25. a term-saturation ranking function. Tell: Are query connectives explicitly composed?

References¶

[1] Gerard Salton, Edward A. Fox, and Harry Wu, “Extended Boolean Information Retrieval”, Communications of the ACM 26(11) (1983), 1022–1036. registry ↩a ↩b Show verification details

[2] Gerard Salton, “The Use of Extended Boolean Logic in Information Retrieval”, SIGMOD Record 14(2) (1984), 277–285. registry ↩ Show verification details