Data collection
Data collection is the process of systematically acquiring or gathering information, observations, or measurements from one or more sources for a defined purpose.
Data for All
Data collection is the systematic gathering of information from actions, observations, systems, or other sources.
- How it is used
- Positions collection as the input that enables organizations to combine records, predict behavior, and target offers.
- How deep it goes
- Explains sources, combinations, incentives, and consequences without specifying collection procedures or technical implementations.
- Examples
- Uses location, browsing, and transaction records, a targeted offer, and the proposed Google–Fitbit combination.
- What you can do after
- Evaluate/understand it by questioning what organizations collect, how they combine it, and whose interests the results serve.
Data-Driven Cybersecurity
Data collection is the gathering of security-relevant data from sources such as logs, network traffic, endpoint activity, and cloud APIs.
- How it is used
- Supplies raw evidence for continuous threat monitoring and incident investigation.
- How deep it goes
- Outlines scoping, sensor deployment, and downstream analysis, but does not detail sensor configuration.
- Examples
- Uses a financial institution breach in which logs, traffic, and alerts helped trace exfiltration and reconstruct an attack timeline.
- What you can do after
- Operate/configure it by defining monitoring scope, deploying collection mechanisms, and supplying evidence for analysis and reporting.
Read how Data-Driven Cybersecurity explains data collection →
Effective Data Analysis
Data collection is the systematic process of gathering, capturing, and recording observations, measurements, or other information for storage or analysis.
- How it is used
- Guides analysts and stakeholders from concept definition to selecting and evaluating data for measurement and reporting.
- How deep it goes
- Defines direct and indirect methods, then gives practical checks for source, refresh, record meaning, change tracking, and quality.
- Examples
- Uses height, historical archives, sales transactions, software clicks, doctor visits, and page-view records to distinguish measures and record meanings.
- What you can do after
- Evaluate data sources for behavioral measurement and choose collection methods aligned with measurable indicators.
Deep Learning with R, Third Edition
Dataset collection is the process of obtaining data instances for use in analysis or machine learning.
- How it is used
- Guides the acquisition and improvement of training data after defining a machine learning task, its inputs, and its targets.
- How deep it goes
- Explains collection choices and curation checks, but stops short of detailed labeling protocols or collection-system implementation.
- Examples
- Uses image classification, spam detection, recommendation, click-through prediction, cookie inspection, and satellite imagery scenarios.
- What you can do after
- Evaluate dataset adequacy and prioritize collection, labeling, cleaning, and feature improvements before model refinement.
Read how Deep Learning with R, Third Edition explains data collection →
Deep Learning with Python, Third Edition
The process of gathering data examples for a machine learning project before model development.
- How it is used
- Guides dataset planning between task definition and model development, including sourcing, labeling, reuse, and collection investment.
- How deep it goes
- Explains collection choices and dataset quality factors through project examples, without detailing operational labeling protocols.
- Examples
- Uses image search, spam detection, recommendations, advertising, cookie inspection, and satellite imagery projects.
- What you can do after
- Apply it by choosing between reusing records, acquiring separate data, collecting new examples, and investing in labeling.
Read how Deep Learning with Python, Third Edition explains data collection →
D3.js in Action, Third Edition
Data gathering is the process of collecting information from one or more sources for analysis or visualization.
- How it is used
- Uses data collection as the foundation for custom visualizations, beginning with desired information and ending with prepared datasets.
- How deep it goes
- Explains source discovery and preparation rationale, but stops short of detailed cleaning and formatting procedures.
- Examples
- Uses Van Gogh paintings, drawings, and letters, plus JavaScript extraction from Wikipedia tables.
- What you can do after
- Apply it via JavaScript to extract painting records from Wikipedia tables for visualization.
Read how D3.js in Action, Third Edition explains data collection →
Grokking Statistics
Data collection is the systematic process of obtaining observations, measurements, or records for a defined purpose.
- How it is used
- Frames data gathering as a question-led step in scientific inquiry, contrasting it with unfocused data mining.
- How deep it goes
- Explains how objectives shape collection and flags bias, missing context, and inaccurate records, without detailing sampling procedures.
- Examples
- Uses family photographs, golf observations, scorecards, and 3D-print filament measurements to illustrate focused collection.
- What you can do after
- Evaluate whether collected observations fit a defined objective and identify context, bias, and accuracy problems before drawing conclusions.
Reinforcement Learning from Human Feedback
Data acquisition is the process of obtaining data from external providers or internal collection operations for a defined purpose.
- How it is used
- Uses vendor-supplied human preference data to improve post-training models against defined evaluation goals.
- How deep it goes
- Details iterative vendor collection, quality measurement, batch refinement, and waste, but does not specify annotation protocols or optimization equations.
- Examples
- Uses a multibatch vendor cycle, ramp-up period, approximately $500,000 contract, and discarded early batches.
- What you can do after
- Operate an iterative preference-data sourcing process that tests quality and model impact before scaling collection.
Read how Reinforcement Learning from Human Feedback explains data collection →
Machine Learning Platform Engineering
Data collection is the process of acquiring data relevant to a defined modeling problem and its target domain.
- How it is used
- Supports modeling after problem definition and KPI selection, while establishing inputs for later ML stages.
- How deep it goes
- Explains collection criteria and lineage rationale, but stops short of prescribing a detailed acquisition workflow.
- Examples
- Uses open datasets, enterprise datasets built from scratch, deployment-domain data, MIDV-500 visuals, and CVAT annotations.
- What you can do after
- Evaluate datasets against relevance, quality, bias, leakage, representativeness, diversity, and lineage requirements.
Read how Machine Learning Platform Engineering explains data collection →
Demand Forecasting Best Practices
Data collection is the systematic gathering of observations or records for analysis or modeling.
- How it is used
- Aligns demand records with forecasting objectives, operational granularity, inventory deployment, and available demand drivers.
- How deep it goes
- Explains what to collect, how to align it, and how to prioritize cleaning, but stops short of prescribing collection systems or algorithms.
- Examples
- Uses weekly warehouse demand, monthly market forecasts, Excel-held data, external providers, and temperature-sensitive ice-cream demand.
- What you can do after
- Operate a forecasting process by gathering relevant demand data, matching its granularity to decisions, and prioritizing feasible improvements.
Read how Demand Forecasting Best Practices explains data collection →
Related concepts
- targeted advertisinguses · Targeted advertising is the delivery of promotional messages to selected audiences based on information about their characteristics, interests, or behavior.
- penetration testingpart of · Penetration testing is an authorized security assessment that simulates attacks against systems to identify and attempt to exploit security weaknesses.
- data sourceuses · A data source is an origin, system, repository, or process from which data is obtained for use or processing.
- preprocessingpart of · Preprocessing is the preparation or transformation of data, content, or other inputs into a form suitable for a subsequent computational operation.
- social mediaused by · Social media is a class of online services that enables users to create, share, discover, and interact with content and with one another.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.