Data Glossary
Browse 103 data terms defined in plain English, from the cultural dictionary of computing.
103 Data Terms
- Ad Tech
- The ecosystem of software, infrastructure, and data systems used to buy, sell, target, measure, and optimize digital advertising. It includes exchanges,...
- Apache Arrow
- A cross-language columnar memory format designed for efficient analytical processing. Unlike Parquet (on-disk), Arrow defines how data lives in RAM, enabling...
- Apache Iceberg
- An open table format for large analytic datasets that brings ACID transactions, schema evolution, time travel, and partition evolution to data lakes. Works...
- Apache Kafka
- A distributed event streaming platform that handles high-throughput, fault-tolerant, real-time data feeds. Messages are organized into topics, partitioned for...
- Apache Kafka
- A distributed event-streaming platform used for high-throughput messaging, logs, and data pipelines.
- Apache Parquet
- A columnar storage file format optimized for analytics workloads. Stores data by column rather than by row, enabling efficient compression (similar values...
- Backfill
- Re-processing historical data through a pipeline to fill gaps caused by bugs, schema changes, new features, or late-arriving data. Requires idempotent...
- Benchmark Contamination
- When benchmark test data leaks into a model's training set, inflating evaluation scores without genuine capability improvement. A pervasive problem in LLM...
- Bias
- Systematic errors in an AI model's outputs that reflect prejudices in training data, algorithm design, or deployment context. An AI trained on biased data...
- Big Data
- A term for datasets or data-processing problems large or complex enough to require specialized tools, distributed systems, or new workflows. In culture, big...
- Binary
- As a noun, a compiled, executable file containing machine code. As an adjective, relating to the base-2 numeral system (0 and 1) that underpins all digital...
- bitmask
- A pattern of bits used with bitwise operations (AND, OR, XOR) to manipulate or test specific bits within a value. Bitmasks are used for permission systems...
- Business Intelligence
- The tools and practices used to analyze company data for operational, financial, and strategic decisions, often through dashboards, reports, and ad hoc...
- Chart
- A visual representation of data such as a line graph, bar chart, area chart, or pie chart used to communicate patterns, comparisons, or trends. In software...
- Control Vector
- A data or representation concept centered on control vector and how information is prepared or encoded for models. It influences how models are trained,...
- Copy
- To duplicate data, files, objects, or values so the same content exists in another location or memory region. In software discussions, copy can also imply...
- Corpus
- A data or representation concept centered on corpus and how information is prepared or encoded for models. It influences how models are trained, evaluated, or...
- Custom Serializer
- A serializer implemented to convert data into a specific wire format, storage format, or schema with rules beyond the defaults. Custom serializers are useful...
- Dark Data
- Data that an organization collects, stores, or generates but does not actively use, understand, or manage well. In data culture dark data is often seen as both...
- Data Access Layer
- A Data Access Layer is a development concept used to shape how software is built, executed, or maintained in real systems. Developers rely on it to make code...
- Data Access Object
- A Data Access Object is a development concept used to shape how software is built, executed, or maintained in real systems. Developers rely on it to make code...
- Data Annotation
- A data or representation concept centered on data annotation and how information is prepared or encoded for models. It influences how models are trained,...
- Data Augmentation Detail
- A data or representation concept centered on data augmentation detail and how information is prepared or encoded for models. It influences how models are...
- Data Binding
- A Data Binding is a development concept used to shape how software is built, executed, or maintained in real systems. Developers rely on it to make code easier...
- Data Cleaning ML
- A data or representation concept centered on data cleaning ml and how information is prepared or encoded for models. It influences how models are trained,...
- Data Contract
- A formal agreement between data producers and consumers that defines the schema, semantics, quality expectations, SLAs, and ownership of a data asset. Like API...
- Data Culture
- The norms and behaviors an organization has around collecting, sharing, interpreting, and acting on data. Strong data culture values evidence and literacy...
- Data Curation
- The careful selection, cleaning, labeling, and organization of data so it is suitable for training, evaluation, or retrieval. Data curation is often one of the...
- Data Drift
- A data or representation concept centered on data drift and how information is prepared or encoded for models. It influences how models are trained, evaluated,...
- Data Driven
- Describing decisions or processes that are strongly informed by data, metrics, or analysis rather than intuition alone. In tech culture the phrase is often...
- Data Ethics
- The ethical questions surrounding how data is collected, stored, used, shared, and interpreted, including privacy, consent, fairness, and power. In modern...
- Data Flow
- A Data Flow is a development concept used to shape how software is built, executed, or maintained in real systems. Developers rely on it to make code easier to...
- Data Flywheel
- A data or representation concept centered on data flywheel and how information is prepared or encoded for models. It influences how models are trained,...
- Data Governance
- The policies, roles, processes, and controls used to manage data quality, access, lineage, retention, and compliance. In data-heavy organizations governance is...
- data gravity
- The principle that large datasets attract applications, services, and additional data to their location, much like physical mass attracts objects through...
- Data Labeling
- A data or representation concept centered on data labeling and how information is prepared or encoded for models. It influences how models are trained,...
- Data Lakehouse
- An architecture that combines the cheap storage and flexibility of data lakes with the ACID transactions, schema enforcement, and performance of data...
- Data Layer
- A Data Layer is a development concept used to shape how software is built, executed, or maintained in real systems. Developers rely on it to make code easier...
- Data Lineage
- The tracking of data's journey from source through transformations to final consumption — which systems produced it, what transformations were applied, and who...
- Data Literacy
- The ability to read, interpret, question, and use data meaningfully without being misled by charts, dashboards, or metrics. In workplace culture, data literacy...
- Data Loader
- A Data Loader is a development concept used to shape how software is built, executed, or maintained in real systems. Developers rely on it to make code easier...
- Data Mapper
- A Data Mapper is a development concept used to shape how software is built, executed, or maintained in real systems. Developers rely on it to make code easier...
- Data Model
- A Data Model is a development concept used to shape how software is built, executed, or maintained in real systems. Developers rely on it to make code easier...
- Data Oriented Design
- A Data Oriented Design is a development concept used to shape how software is built, executed, or maintained in real systems. Developers rely on it to make...
- Data Parallelism ML
- A data or representation concept centered on data parallelism ml and how information is prepared or encoded for models. It influences how models are trained,...
- Data Quality
- The degree to which data is accurate, complete, relevant, consistent, and usable for a specific AI task. Poor data quality often causes more serious problems...
- Data Race
- A Data Race is a development concept used to shape how software is built, executed, or maintained in real systems. Developers rely on it to make code easier to...
- Data Scaling
- Increasing the amount or effective use of training or retrieval data to improve model behavior, often in line with scaling-law considerations. Data scaling is...
- Data Science
- A field combining statistics, programming, domain knowledge, and data analysis to extract insight, build models, or support decisions from data. In tech...
- Data Serialization
- A Data Serialization is a development concept used to shape how software is built, executed, or maintained in real systems. Developers rely on it to make code...
- Dataset Bias
- A data or representation concept centered on dataset bias and how information is prepared or encoded for models. It influences how models are trained,...
- Dataset Card
- A data or representation concept centered on dataset card and how information is prepared or encoded for models. It influences how models are trained,...
- Dataset Distillation
- A data or representation concept centered on dataset distillation and how information is prepared or encoded for models. It influences how models are trained,...
- Dataset Versioning
- A data or representation concept centered on dataset versioning and how information is prepared or encoded for models. It influences how models are trained,...
- Data Store
- A Data Store is a development concept used to shape how software is built, executed, or maintained in real systems. Developers rely on it to make code easier...
- Data Transfer Object
- A Data Transfer Object is a development concept used to shape how software is built, executed, or maintained in real systems. Developers rely on it to make...
- Data Type
- A Data Type is a development concept used to shape how software is built, executed, or maintained in real systems. Developers rely on it to make code easier to...
- Data Validation
- A Data Validation is a development concept used to shape how software is built, executed, or maintained in real systems. Developers rely on it to make code...
- Data Visualization
- The visual presentation of data through charts, graphs, maps, and other forms intended to make patterns easier to understand. In culture, good data...
- dbt
- Data Build Tool — a transformation framework that lets data analysts write modular SQL to transform raw data into analytics-ready tables. Models are...
- Deserialization
- A Deserialization is a development concept used to shape how software is built, executed, or maintained in real systems. Developers rely on it to make code...
- EHR
- Electronic Health Record — a digital version of a patient's medical history maintained by healthcare providers. EHR systems (Epic, Cerner) are notoriously...
- ETL
- Extract, Transform, Load — the three steps of moving data between systems. Extract from sources, transform into the desired format, load into the destination....
- Event Sourcing
- An architectural pattern where state changes are stored as an immutable sequence of events rather than just the current state. The current state is derived by...
- Export Culture
- A team or product habit where users repeatedly export data to spreadsheets, CSVs, or other tools because the core system does not support the analysis or...
- Face Embedding
- A data or representation concept centered on face embedding and how information is prepared or encoded for models. It influences how models are trained,...
- Fake Data
- Artificial or simulated data used for testing, demos, prototyping, or development instead of real production data. In engineering culture fake data is useful,...
- Graph Culture
- A team habit of thinking heavily in terms of graphs, dependencies, relationships, and network-like structures, whether in data, architecture, or social...
- Idempotent Pipeline
- A data pipeline that produces the same output regardless of how many times it's run with the same input. Achieved through techniques like UPSERT/MERGE,...
- Immutable
- Data that cannot be changed after creation. Instead of modifying existing data, you create new copies with the changes applied. A cornerstone of functional...
- Immutable Data Structure
- An Immutable Data Structure is a development concept used to shape how software is built, executed, or maintained in real systems. Developers rely on it to...
- Labeled Data
- A data or representation concept centered on labeled data and how information is prepared or encoded for models. It influences how models are trained,...
- Label Smoothing
- A data or representation concept centered on label smoothing and how information is prepared or encoded for models. It influences how models are trained,...
- Machine Learning
- Machine learning is a subset of artificial intelligence in which systems learn patterns from data rather than being explicitly programmed with rules. Instead...
- Medallion Architecture
- A data organization pattern with three layers: Bronze (raw ingestion, no transformation), Silver (cleaned, deduplicated, conformed), and Gold (aggregated,...
- Metadata
- Data that describes other data — such as a file's creation date, an image's resolution and EXIF tags, a database table's schema, or an API response's...
- Migrate
- To move code, data, traffic, or users from one system, version, or platform to another in a controlled way. In practice the word often implies more hidden risk...
- NoSQL
- A category of databases that don't use traditional relational tables and SQL — including document stores (MongoDB), key-value stores (Redis), column stores...
- Object
- Object in programming is a fundamental concept in object-oriented programming (OOP) representing a self-contained unit that bundles related data (called...
- OpenStreetMap
- A collaboratively maintained open geographic database used in mapping applications and spatial tools.
- oracle (blockchain)
- A service that feeds external real-world data (prices, weather, sports scores, API responses) into a blockchain, since smart contracts cannot natively access...
- Pipeline
- A series of automated steps that process data or code sequentially. In CI/CD, a pipeline builds, tests, and deploys code. In data engineering, a pipeline...
- Property
- A named attribute of an object or class that stores data or computes a value. In many languages, properties use getter/setter methods behind a field-like...
- Query
- A request for data from a database or search engine, typically expressed in a structured language like SQL, GraphQL, or a search DSL. As a verb, to query means...
- Schema
- A formal description of the structure of data — including tables, columns, types, constraints, and relationships in a database, or the shape of JSON, XML, or...
- Schema Registry
- A centralized service for managing and enforcing schemas for Kafka messages (Avro, Protobuf, JSON Schema). Producers register schemas before sending messages;...
- Self-Instruct
- A method of generating synthetic instruction-following data using models themselves, often to expand training datasets without fully manual labeling....
- Series
- A one-dimensional labeled array in data libraries like pandas, capable of holding any data type and supporting vectorized operations, alignment by index, and...
- Slowly Changing Dimension
- A data warehousing technique for tracking how dimension attributes change over time. Type 1 overwrites the old value, Type 2 adds a new row with version dates...
- Star Schema
- A dimensional modeling pattern where a central fact table (events, transactions) connects to multiple dimension tables (who, what, when, where) via foreign...
- State
- The current set of data values that determine a program's behavior and UI at any given moment, managed locally in components, globally in stores, or on the...
- Stream
- A sequence of data elements made available over time, processed incrementally rather than loaded entirely into memory, used for file I/O, network...
- Stream Processing
- Processing data continuously as it arrives rather than in periodic batches. Handles unbounded datasets using windowing (tumbling, sliding, session) for...
- Sync
- Short for synchronous — describes an operation that blocks the calling thread until it completes and returns a result, as opposed to async operations that...
- synthetic training data
- Training data generated by AI models rather than collected from humans, used to train or fine-tune other models. While dramatically cheaper to produce,...
- Table
- A structured collection of data organized in rows and columns within a relational database, where each row represents a record and each column represents a...
- Time Series
- A sequence of data points indexed by timestamp, typically collected at regular intervals — used for metrics, sensor readings, and financial data, and stored in...
- Timestamp
- A value representing a specific point in time, typically stored as seconds or milliseconds since the Unix epoch (January 1, 1970 UTC) or as an ISO 8601 string...
- Training Data
- The dataset used to teach a machine learning model. The quality, size, and representativeness of training data fundamentally determine a model's capabilities...
- Training Data Licensing
- The legal and contractual framework governing whether data can be used to train or fine-tune AI models.
- Type Provider
- An F# compiler feature that generates types at compile time from external data sources (databases, APIs, CSV files). Gives you IntelliSense and type safety for...
- Unstructured
- Lacking a predefined schema, formal format, or consistent organization. The term appears in data work, writing, communication, and sometimes in criticism of...
- Unstructured Data AI
- AI systems or workflows focused on processing unstructured data such as documents, images, conversations, audio, or free-form text rather than neatly labeled...
Related Topics
- Architecture (21 terms in common)
- Backend (17 terms in common)
- Ml (16 terms in common)
- Representations (16 terms in common)
- Ai (6 terms in common)
- Governance (5 terms in common)
- Analytics (5 terms in common)
- Database (4 terms in common)
- Pattern (4 terms in common)
- Sql (3 terms in common)
- Databases (3 terms in common)
- Decision Making (2 terms in common)
- Design (2 terms in common)
- Operations (2 terms in common)
- Streaming (2 terms in common)
- Visualization (2 terms in common)
- Messaging (2 terms in common)
- I/O (2 terms in common)
- Ethics (2 terms in common)
- Apis (2 terms in common)