Welcome to Data Extraction Techniques. After watching this video, you will be able to: List examples of raw data sources. Describe data extraction techniques. Relate use cases with data
sources and extraction techniques. Here are some examples of raw data sources: Archived text and images
from paper documents and PDFs Web pages, including text,
tables, images, and links Analog audio and video, which can be recorded on media such as
magnetic tapes or streaming in real time Survey data, statistical, and economic data Transactional data from business, financial, real
estate, and point-of-sale, or POS, transactions Here are more examples of raw data sources: Event-based data such as social media streams Weather data from weather station networks Internet of Things (or IoT) sensor streams Medical records, such as prescription history,
medical treatments, and medical images Personal genetic data encoded
in DNA and RNA samples Evidently, data is everywhere. And much
of it is highly sensitive and personal, and needs to be very carefully guarded
for privacy and other concerns. There are many techniques for extracting data, depending on the kind of data source and the
intended use of the data. Examples include: Optical character recognition (OCR),
which is used to interpret and digitize text scanned from paper documents so it
can be stored as a computer-readable file Analog-to-digital converters (ADCs), which can
digitize analog audio recordings and signals, and charge-coupled devices (CCDs)
that capture and digitize images Opinions, questionnaires, and vital statistical
data obtained through polling and census methods Cookies, user logs, and other methods used
for tracking human or system behavior More techniques include: Web scraping, used to crawl web pages in
search of text, images, tables, and hyperlinks. APIs, which are readily available
for extracting data from all sorts of online data repositories and feeds, such as
government bureaus of statistics, libraries, weather networks, online
shopping, and social networks. SQL languages for querying relational databases, and NoSQL for querying document, key-value,
graph or other non-structured data repositories. Edge computing devices, such as
video cameras that have built-in processing that can extract features from raw data Biomedical devices, such as microfluidic
arrays that can extract DNA sequences Here are a few high-level examples of use cases, along with their raw data sources
and extraction techniques. You can use APIs to extract
data from multiple structured data sources for integration
into a central repository. You can also use APIs to capture periodic or asynchronous events to store
them in a history archive. Rather than transmitting potentially
very large volumes of redundant data from IoT devices, you can use edge
computing to reduce that data volume by extracting features of
interest from the raw data. Often, this kind of extraction
at the source is impractical so the data is migrated to storage as-is for
further processing, analysis, or modeling. You can use medical imaging devices
and biometric sensors to acquire data for diagnostic purposes. In this video, you learned that: Some examples of raw data sources are archived
text and images from paper documents and PDFs, and web pages, including text,
tables, images, and links. Many extraction techniques rely on sophisticated
technology to capture information from raw data. SQL, NoSQL, web scraping, and APIs are
important techniques for extracting data. You can use medical imaging devices and biometric sensors to acquire
data for diagnostic purposes.