postingsrecord

A postingsrecord, or posting record, is a unit of data used in inverted indexes to describe a single occurrence of a term within a document. It encapsulates the essential information needed to locate and analyze that term during search and retrieval tasks.

A typical postingsrecord includes a document identifier (docID), the term frequency within that document (tf), and may also store positional information indicating where the term appears in the document. In many systems, additional data such as character offsets, field identifiers, or payloads (arbitrary metadata) can be attached. Postings records are often organized into a postings list for a given term, with entries sorted by docID.

Postings records are the backbone of an inverted index. They enable efficient query processing, including exact term matching, phrase queries, and proximity searches. Depending on the index format, postings may be stored in memory or on disk and often use compression or encoding schemes (for example, delta encoding for docIDs and tf, or variable-length integers) to reduce space. In some architectures, there are separate structures for positional data or for offsets and payloads, so a term’s postings can be accessed with different levels of detail.

In practice, a postingsrecord might look like: docID, tf, positions=[p1, p2, ...], offsets=[start1, end1, ...]. The concept is general and used across information-retrieval systems, digital libraries, and search engines as part of the mechanism that maps terms to the documents that contain them.