What Is Unstructured Data Used For?

What Is Unstructured Data Used For?

You may not have heard of it, but data is a big deal. And there is a challenge with all this new information coming at us daily. More than 90% of it is unstructured. But what does that mean and why should you care about this?

With the recent flood of unstructured data, it has become easier than ever for businesses to make use of this new source. The goal of analyzing all types or formats is to discover actionable value.

What Is Unstructured Data?

Unstructured data are data contents that cannot be neatly fitted into a spreadsheet or accessed as easily as a database. It is difficult to store and manage, but creative minds have been finding new ways all the time.

The challenge with unstructured data is that it is not easily stored in a traditional column-row database or spreadsheet like Microsoft Excel. It can be difficult both to analyze and to search, which has made this type of information less useful for organizations. Less useful until recently now that we have tools powered by Artificial Intelligence (AI). These new AI programs were specifically created to access insights from these types of unstructured data sets.

Unstructured Data Importance

As more and more data gets generated, it is important that organizations find ways to manage this information so they can act on the insights. This helps businesses prosper in competitive environments where having Big Data access always translates into success for an organization no matter what their industry might be.

Since most unstructured data contents do not follow any grid-like pattern whatsoever, it takes some serious digging through databases or spending hours wading through countless pages online. Such efforts are needed if you

want something meaningful enough to report back with good results. Thankfully, Machine Learning (ML) has come alongside human intelligence.

Unstructured Data Examples

Unstructured data is any kind of information that does not fit into a spreadsheet or database structure. The most common examples are email or document texts, social media posts, or chat messages. These all lack grid order and can be difficult to analyze because there is not always an obvious way you would categorize them.

Here are some common types of unstructured data:

· Text files

· Photos

· Video files

· Audio files

· Web pages and blog posts

· Social media sites

· Presentations

· Call center transcripts

· Open-ended survey responses


For data-driven organizations, data is the lifeblood of any company and knowing what information to collect can be difficult. Business Intelligence (BI) and analytics tools such as Natural Language Processing (NLP) are being developed that have been specifically created with this goal in mind –analyzing all types or formats so you can discover actionable insights on demand.

The Content Analytics Platform (CAP), developed by Scion Analytics, can help with data discovery. The CAP can quickly analyze any textual data, in any format, and turn it into structured data that is discoverable for value.

What Is Unstructured Data Discovery?

What Is Unstructured Data Discovery?

In the age of big data, it is crucial to be able to identify and provide visibility into where your organization stores its information. This process is known as Data Discovery –identifying structured or unstructured data stored in various databases. These can be found across a variety of different sources (like social media sites). Data volume consumed at any given time.

Unstructured data is information in many different forms that does not follow conventional models. Basically, it is data that cannot be neatly fitted onto a spreadsheet. It is difficult to store and manage, but creative minds have been finding new ways all the time.

With the recent flood of unstructured data, it has become easier than ever for businesses to make use of this new source. Business Analytics (BI) and analytics tools are being developed that have been specifically created with the goal in mind. The goal of analyzing all types or formats to discover actionable value.

Unstructured Data Discovery Importance

Businesses must know what data they hold, its contexts, and where it is located before they can protect that data from compromise, exfiltration, or threats. This holds true for ensuring data accuracy and complying with various privacy and security mandates.

When the data is properly discovered, it can be classified. Data classification is the process of tagging data according to its type, its sensitivity or confidentiality, and the cost if altered or stolen. 

With classification information, you can implement security controls to protect data from accidental or intentional compromise as well as comply with the compliance mandates.

Benefits of Unstructured Data Discovery

When businesses know what data they hold, its contexts, and where it is located, they can then benefit in these ways:

Implement Data Classification Processes

Data classification is an important step in the process of making data easily searchable and trackable. It also eliminates multiple duplications, which can reduce storage costs while speeding up this entire procedure.

Apply Context-Sensitive Security Controls and Access Controls

With Context-Sensitive access, you can decide which apps a user has access to, which are based on their context. For example, if there are certain IT policies in place and they want the users’ devices to conform, then this feature will allow for it.

Execute Context-Sensitive Disaster Recovery Exercises

Context-Sensitive weighs the who, what, when, and where of access requests according to organizational needs. 

Conduct Context-Sensitive Monitoring and Audits

Contextualized security systems empower organizations to take a more holistic approach when it comes to real-time decision making.


Data is the lifeblood of any company and knowing what information to collect can be difficult. However, when you have a discovery process in place, data classification becomes much easier.

In the age of Big Data, it’s easier than ever for businesses to extract value from unstructured data. BI and analytics tools are being developed that have been specifically created with this goal in mind –analyzing all types or formats so you can discover actionable insights on demand.

The Content Analytics Platform (CAP), developed by Scion Analytics, can help with data discovery. The CAP can quickly analyze any textual data, in any format, and turn it into structured data that is discoverable for value. 

Unstructured Data In Government

Unstructured Data In Government

Unstructured data that is gathered by the federal government is becoming a huge challenge to manage for easy accessibility. Data is becoming more and more important in our daily lives. It can be found everywhere, from social media posts to blog entries. All sorts of unstructured data are just waiting for someone who has the know-how on how best use them. But there is one problem: there is not really any good way of organizing this information.

Unstructured data is any electronic file that does not land in a structured platform, such as a database where it is easy to organize and access. Government centers are full of unorganized information collected from outside sources like social media posts or blog entries along with images and videos.

The Air Force collects a lot of data while its planes fly. Sensors record weather information, the aircraft collects GPS mapping data, and onboard systems track operational performance. Usually, this kind of raw material would come in an unstructured form. Different disciplines have their own storage silos.


There are many challenges to harvesting and managing unstructured data efficiently.

Files are collected from outside sources such as social media and blog posts along with images or videos for easy organization to facilitate access. There can be several different ways files might get recorded though. Some agencies will set up their system so they can capture those unstructured categories inside a database, making them easier to sort and search through. Otherwise, the data will be siloed and made inaccessible.

For decades, the amount of unstructured data has exploded. Now, agencies are struggling to keep up with it all. There is a lack efficiency in their management practices because they are not fully utilizing current platforms like Artificial Intelligence (AI) and Machine Learning (ML). These

technologies can help with this information overload by identifying patterns or trends on high volume levels.

The data growth in the datacenter is also being driven by several new initiatives, such as the Internet of Things (IoT) and AI. This spans across the edge-core-cloud.

The need to store large amounts of data led to the evolution from simple storage into more complicated systems. Agencies used stacked appliances and then silos to manage all this new influx –but at what cost? It increased costs while decreasing efficiency.


There are many insights to be gathered from unstructured data using smart practices with the right tools of AI and ML.

Using a data lake, a government health agency could store X-rays so that ML can associate common characteristics of bone degeneration or disease development. This would lead to faster diagnoses because they are less likely to miss rare diseases if there is evidence for it already within their database.

Using the unstructured data of marine biology from vessels’ traversing, the Navy is working on a better coating for its ships. This will keep them from rusting or degrading. They need to get the most effective design and creation models. And they need to understand what marine life these vessels move through with regularity.

Sensors on the farm pick up humidity levels or the pH balance of the soils to enable better planning for suppliers and supply routes.


Although the vast majority of data is unstructured, much insightful benefit can still be gathered from it. It requires human ingenuity using the right processes and the right tools such as Artificial Intelligence (AI) and Machine Learning (ML).

Unstructured Data In Healthcare

Unstructured Data In Healthcare

There are vast amounts of unstructured data in the healthcare industry. It is no surprise that the healthcare industry is struggling with data overload. During every patient encounter, providers rely on hundreds of discrete signals from lab results and qualitative descriptions. These are used to make decisions about treatment plans for patients. These varied, unstructured, discrete signals make it challenging to easily access healthcare data. Clinicians need easy access, but what is the best way to organize all this unstructured clinical intelligence into innovative insights which can improve patient outcomes?

When healthcare practitioners began storing and managing data digitally, they (like people in most industries) used structured data. This type of information can be clearly recorded with specific ways machines understand like a relational database or spreadsheet that contains names, dates, etc. Alphanumeric characters such as those used in ICD-10 Codes and CPT codes are part of this kind of record keeping system.

Electronic Health Record (EHR) systems are designed to use structured data. The Structured Data Capture (SDC) standards initiative is working toward a goal of interoperability so healthcare organizations can share information for patient care.

Unstructured Data In Healthcare Challenges

In order to process unstructured data to be actionable for clinicians, the healthcare industry needs ways of overcoming numerous challenges, listed below:

Interpreting Handwriting

Handwritten notes, which are a gold mine of information written by patients, nurses, or physicians, have never been easily accessible as unstructured information.

Reading Radiological Images

Reviewing and analyzing X-ray, CAT, MRI, or ultrasound medical images requires the skill of experienced professionals. These professionals may need to access past procedures, stored in unstructured form, to distinguish healthy tissue from abnormalities.

Creating Metadata

There is a way to pre-process unstructured data for use in an EHR or other system that requires structured data. For example, technicians and physicians can describe images using codes and keywords which are then entered into the computing system. This is done by adding information about what was seen through this metadata making them searchable.

Accommodating File Sizes

Unstructured data generally requires more storage than structured. Unstructured information is measured by terabytes and petabytes, not gigabytes. This requires greater capacity storage infrastructure.

Performing Searches

Without metadata, many forms of unstructured information would be impossible to search (e.g., images, or audio from a conference).

Using Streaming Data

Data is becoming more and more dynamic. The Internet of Things (IoT), edge computing, and Artificial Intelligence (AI) are all able to analyze this new type of information in real time.

Managing Massive Data Volumes

Data overload is the bane of modern healthcare professionals. They will need applications that help them prioritize and manage all this new, unstructured data.

Dealing With Existing Records

There are huge stores of unstructured healthcare data that could potentially aid better treatment decisions. These data are inaccessible to the EHR systems. AI and ML are changing the way healthcare providers store, analyze, and share data. These technologies are making it possible to have a more efficient patient care experience now by utilizing unstructured information that was previously inaccessible in EHRs. These will be needed tomorrow too as we continue moving forward with time-sensitive health initiatives. A good example is personalized medicine or precision oncology treatments, which require ever increasing amounts of precise knowledge.

Unstructured Data Examples

Unstructured Data Examples

The various kinds of data that are generated through a variety of sources at practically every moment are categorized as structured, unstructured, or semi-structured data. Unstructured data is usually complex and heterogeneous; it cannot be mapped to a predefined structure such as a data table or a relational database. There are many types of data, each are different and has its own importance.

Here are eight examples of unstructured data:

Healthcare Records

Healthcare generates large volumes of machine-generated, as well as human-generated, unstructured data. Medical imaging devices such as endoscopes, laparoscopes, and surgery robots generate a plethora of this type. Patient monitors in operating theaters or intensive care units may provide some biosignals for analysis.

Social Media

Social media has become an integral part of the lives of billions of people around the world, with people using them to consume information and interact socially. Businesses will often use social platforms as well to reach their customers more effectively than ever before. Governments have also started relying heavily on this type of communication tool when they need public opinion polls quickly. Disaster victims use it for assistance when managing crises at home or abroad from afar.

Business Documents

The plethora of information that is used to conduct business, such as emails and presentations, contain data in the form of text messages. These documents are unstructured, which makes them difficult for organizations

because they are not mapped onto an already existing database system like Excel or Access.

Multi-Media Content

The media and entertainment industry are always creating new content, whether it is for surveillance systems or professional publishers. However, these databases do not process the actual contents of video files because they are stored unstructured data with images as tags instead.

Electronic Communications

For today’s professional and personal discussions, data can be created in the form of unstructured audio or text by using popular apps such as WhatsApp. Web conferencing platforms like Zoom offer additional options for collaboration tools that are used to communicate across distances.


There are many methods for conducting market research and employee engagement. The format of such surveys typically includes multiple choice questions, with open-ended responses being formatted as unstructured text.

Publications And Listings

In addition to the structured data published on sites like LinkedIn, publications are also filled with unstructured content. This is hard to extract for machines and computer users alike. This includes news stories from newspapers or magazines about current events as well as job listings.


Websites are made up of text, images, audio, and videos that can be found on the web. All this content has an output tone that is not captured in HTML codes.


Data is the lifeblood of any company and knowing what information to collect can be difficult. Unstructured data is information in many different forms that does not follow conventional models. Basically, it is data that cannot be neatly fitted onto a spreadsheet. It is difficult to store and manage, but creative minds have been finding new ways all the time.

The Content Analytics Platform (CAP), developed by Scion Analytics, can help with data discovery. The CAP can quickly analyze any textual data, in any format, and turn it into structured data that is discoverable for value.

Unstructured Data Myths

Unstructured Data Myths

Unstructured Data Analytics (UDA) is one of the newest and hottest subfields in Big Data today. However, it comes with its fair share of misconceptions. This either causes businesses to be too hesitant to buy into this technology or to misinterpret what UDA can do for them. This in turn causes them to fail using it efficiently. UDA is analysis of unorganized data such as a string of text or numbers. So, let us clear up exactly what UDA does and how helpful are they?

Here are some common myths about UDA plus explanations regarding their potential use cases:

It Is Only Textual Data

Data analytics is changing the way we look at data. Data is everywhere, and it’s not just text anymore. Even photographs, videos, facial expressions can all be turned into numbers that tell a story about us. Such stories as our preferences for movies or clothes, how much time we spend on social media sites, or if someone makes us feel happy. These insights may even help companies decide which products to develop next.

Merrill Lynch estimates that over 80 percent of data found today is unstructured. But not all this information is structured as text alone. Unstructured formats include IMS chat reviews, and customer communication. It also includes scientific measurements such as seismographs and more ordinary forms such as body language or tone of voice.

It Is Not Helpful To Companies Without Chat Or Customer Review Data

Companies that don’t rely on customer reviews or chats to understand what their clients are thinking might be lacking service. They might not have enough content to analyze, or simply aren’t in the field of business. However,

any enterprise that communicates stands to benefit from UDA because it can help figure out how employees need support and address company image issues. Sometimes the most important part of repairing company-client relations is starting within your own organization by using UDA on internal emails for example.

It Only Means Sentiment Analysis

A common misconception is that UDA works only on Sentiment Analysis. Sentiment Analysis is a feature that works within advanced UDA by measuring how people feel about something (positively or negatively). It analyzes many different features of the data to construct a big picture including syntactical analysis (to determine importance based on syntax), topic categorization (by theme), and geospatial organization (which uses geography and setting to aid in the interpretation of the data). Advanced UDA platforms work with many other features to construct the big picture of data, not just sentiment analysis.

Taxonomies Are Enough To Conduct Unstructured Data Analytics

Taxonomies are no longer enough to organize today’s data. Taxonomies can sort the information into specific categories like “long” or “short,” but it cannot go further and do what contemporary UDA initiatives do: extract meaning from different types of candy boxes without ever tasting them.


The big data that is available is mostly unstructured data. But that data is still potential revenue for data-driven businesses. Advanced software technologies empowered with Artificial Intelligence (AI) and Natural Language Processing (NLP) can help. The Content Analytics Platform (CAP), developed by Scion Analytics, can do this by quickly rearranging any unstructured textual data into highly usable structured data.

Data Mining Vs. Process Mining

Data Mining Vs. Process Mining

Data mining is a part of Business Intelligence (BI) that seeks to understand relationships and patterns in large datasets found in big data. Big data is a term referring to massive databases of both structured and unstructured static information that can be exploited for business intelligence needs. This data can have the potential for improving business operations. The data is mined from such sources as emails, data storage, phones, applications, and databases. This data is mined, processed, and analyzed. Companies gain insights from these harvested data.

Process mining seeks to understand real-time procedure steps to detect inefficiencies or make improvements in the accomplishing of a business task. Process mining is the analyzing and monitoring of business processes. Data is gathered through, or mined from, corporate information systems which displays the actual process. It does this by capturing a time-stamp and an event log of each of the process steps. The process mining is accomplished by using strong algorithms combined with advanced data transformation enabling the discovery and improvement of the business processes.


There are similarities between data mining and process mining. Both are a subset of business intelligence (BI) and both access large volumes of data to achieve information for action. Both use algorithms to obtain hidden patterns and relationships within the data.


Data Mining Finds Static Data

Data mining is static and used by corporations to analyze big datasets to predict business patterns. The data analyzed is harvested from static datasets such as databases, which are available records. It looks for things like what group of consumers will buy what product, or where does a marketing effort have the greatest impact. Data mining has no concern with business processes.

Process Mining Finds Dynamic Data

Process mining is dynamic and gathers needed information from created actions. It can be from real-time events provided through a live feed. It looks for steps that are inefficient or time-consuming to control and improve those steps. It reveals a true, end-to-end process.

Data Mining Looks At Arbitrary Data

Data mining obtains information from what happens to be available. Data mining arbitrarily gains information from large databases without targeting a specific inquiry.

Process Mining Looks At Real-Time Data

Process mining targets a specific question about a process. Process mining gets current activity.

Data Mining Looks At Results

Data mining can only look at the results of available data. It cannot answer how those data came to be.

Process Mining Looks At Causes

Process mining can see the cause of actions.

Data Mining Analyzes Patterns

Mainstream patterns are analyzed by data mining. Exceptions to those mainstream patterns are not considered for analysis.

Process Mining Sees Exceptions

But exceptions and irregularities can be very useful for the process mining technique. They could provide clues to what is not working well and what needs improvement.


Both data mining and process mining serve important purposes in the realm of business intelligence (BI). They are necessary for successful, efficient business operations. Data mining provides the source of market knowledge for companies to make smart decisions. The analyzed results are applied in various industries such as retail, journalism, and scientific research. But process mining provides the knowledge of operations that help companies improve and function smartly.

Disadvantages Of Unstructured Data

Disadvantages Of Unstructured Data

Difficult To Analyze

Since there is a lot more unstructured data than structured data, there are also challenges associated with this type of data. First, unstructured data is difficult to analyze. An average person with a working knowledge of Excel cannot mine unstructured data. This is more realistic for using structured data sets with business intelligence. Processing unstructured data is reserved for data scientists and data analysts with proper training and tools. Unstructured data in raw format is difficult to wrangle and interpret.

Requires Specialized Tools

Second, unstructured data requires specialized tool. Most businesses invest in a specific data management tool to analyze data. For example, the text analytics platform from Scion Analytics, the CAP liberates value from unstructured data with AI and NLP powered capabilities. With minimal training, a business can use the CAP to analyze and process data they did not even know they had. These new insights could amount to data drive an actionable decisions on business intelligence that was previously based on guesswork.


Other challenges of unstructured data include the storage aspect. Structured data has a predefined format, and it is easy to store and organize. Due to a lack of schema and structure, unstructured data is expensive and difficult to store. Having to manage the storage aspect of unstructured data is just one facet that differentiates it from structured data.

Indexing Difficulties

Since structured data has been traditionally used for a long time, approaches to unstructured data are still being developed. Indexing unstructured data is difficult and prone to error due to free form structure and a lack of pre-defined attributes. The difficulty of accessing and analyzing unstructured data makes search results not very accurate. Finally, the complexity of unstructured data makes security a challenge. Whereas security methods for structured data are available, they are still being developed around the security of unstructured data.


The buzz about unstructured data has gripped businesses across industries. Many businesses from healthcare to technology have been revolutionized by access to previously hidden insights that reside in unstructured data. Executives across the world embraced unstructured data with its boundless opportunity.

A new crop of technological advances grew up around unstructured data. Text analytics platforms, statistical models, and AI powered tools have been developed to harness unstructured data. This innovation has not gone unnoticed by the business sector. Benefits of unstructured data are far reaching in implications. For a business if 80-90% of data is unstructured, the insights gained from analyzing this data set are unlimited. These insights create new revenue streams, scale businesses, and push innovation forward into the future.

What Is Semi-Structured Data?

What Is Semi-Structured Data?

Semi-structured is made up of partially unstructured data and partially data structure created by metadata. It is an interesting intersection between the two data types and it can yield transformational insights when analyzed. A good example of semi-structured data is an X-ray. An X-ray consists of a great many pixels. The sheer volume of pixels cannot be searched, queried, and analyzed like structured data. However, X-rays, like most files, contain metadata. This “data about data” is what enables semi-structured and unstructured data to be harnessed.

Every day data is used to shape the direction of businesses, develop new business offerings, and gain a competitive advantage. In a market where businesses pursue innovation and thrive off disruption, harnessing data is integral to success.

Data is everywhere. It comes in different shapes and sizes. There are three different types of data: structured data, unstructured data, and semi-structured data.

Structured data is what has traditionally been thought of as data. It exists in predefined formats and is easily accessible by the average user. Unstructured data is the new frontier of data. It lacks a predefined format akin to endless chaos of data full of powerful insights. The middle between structured data and unstructured data is semi-structured data.

Semi-Structured Data Examples

An X-ray is just one example of unstructured data. Upon further examination, numerous examples of unstructured data can be found in everyday business operations.


Email is a great example of semi-structured data. The popularity of the Internet and the proliferation of social media has created a deluge of new data. This data comes in flexible formats collected from a vast population sample. Digital communications such as emails are a source of semi-structured data. This is because every email has:

– Subject

– “To” line

– “From” line

– Date stamp

– Time stamp

The above fields are sources of structured information. However, the text body of an email is a form of unstructured data. There is no defined format or character limit for the body of an email. Emails collected, searched, and analyzed across an enterprise can represent a powerful source of information and record-keeping. Furthermore, emails can provide data mining opportunities to:

– Analyze customer feedback

– Streamline customer support

– Target marketing initiatives

– Develop social media initiatives

– Shape strategic initiatives

Web Pages

The overwhelming popularity of the Internet has produced a lot of content. Another example of semi-structured data is web pages. Most web pages have an organization with tabs for:

– Home

– About Us

– Blog

– Services

– Contact

These tabs are easy to navigate and search thus representing structured data. However, the web pages are written in HTML containing text and data within each of these pages that have no structure. A wealth of information lies hidden in web pages across the internet. Once a business knows how to leverage digital resources, both internal and external, it will become more successful.

Unstructured Data Vs. Semi-Structured Data

The availability of semi-structured data poses the question: what is the difference between semi-structured and unstructured data? It is a grey area that leaves a lot open to interpretation. All documents, images, and other files have some form of data structure. Therefore, it is hard to distinguish where semi-structured data ends and unstructured data begins. Both semi-structured and unstructured data lack organization and rules that are present in a relational database of structured data.


As more technological advances of data analytics tools evolve, the understanding of semi-structured data and how it relates to unstructured and structured data will deepen. For now, semi-structured data remains a prolific presence on the internet capable of taking businesses forward into the future.

Pros And Cons Of Structured Data Pt:2

Pros And Cons Of Structured Data Pt:2

In today’s competitive market, technological advances are evolving at the speed of business. For a business to be competitive, it needs to innovate and change with the times. One of the most profound, recent developments has been the use of unstructured data.

Data is integral to any business. Historically, data can be of any size and shape classified into structured data and unstructured data. Most businesses have tapped into structured data and its benefits. It has a predefined format and structure that makes it easy to access for business intelligence. Examples of structured data include credit card information and Excel files.

Conversely, unstructured data is more difficult to access and analyze. It can be thought of as being in a chaotic state, an endless alphabet soup of data with hidden insights. Examples of unstructured data include social media posts and emails.

Businesses have been using structured data techniques for some time while unstructured data applications remain to be explored. As more applications are found for unstructured data, data analytics tool will evolve as well. Because unstructured data is still in nascent stages, there are distinct pros and cons to using one type of data over the other.

Pros of Structured Data

1. Ease Of Use

One of the benefits of structured data is how easily it can be used by the average user with a working understanding of the data. It enables self-service of structured data without an in-depth understanding of the complexities of the data. It must be noted that structured data is also easy to use by machine learning. The organized nature of unstructured data lends itself to analysis and manipulation.

2. Convenient Storage

Due to its predefined structure, structured data is conveniently stored in data warehouses. Data warehouses are optimized to save storage space for enterprises and to encourage easy data access. Conversely, unstructured data which is much less defined is stored in data lakes with much greater storage capacity.

3. Access To More Tools

Historically, structured data was the only option for businesses looking to quantify data. Data analytics tools and practices have been developed around structured data. While unstructured data is still in its infancy, data managers have efficient tools at their disposal to process structured data.

Cons Of Structured Data

1. Limitations On Use

Structured data is predefined in its format lending itself to be used for intended purposes. This organizational structure places limitations on its flexibility and use cases. The opposite is the case for unstructured data which has a free form format and can be repurposed for multiple use cases.

2. Limited Storage Options

Structured data is stored in data warehouses. Data warehouses have rigid rules about data storage. Any change to structured data is labor intensive requiring a lot of resources and time to update. Some businesses use cloud-based data warehouse to eliminate the need for on-prem equipment and increase scalability.


Traditionally, businesses have used structured data to gain greater insight into operations. Due to its longevity, structured data has better tools, convenient storage and is easy to use. As the predecessor to unstructured data, structured data remains the reliable and preferred source for data analytics.