|
Cataphora C-Analytics
Midway upon the journey of our life
I found myself within a forest dark,
For the straightforward pathway had been lost
- Dante, Inferno Canto 01
What is an Analytic?
An analytic can be described as a mapping of logical inferences drawn from potentially numerous and disparate sources of fact. Dante couldn't find his pathway because the forest trees were in the way - metaphorically speaking - and thus he could make no spatial sense of his surroundings. And - metaphorically speaking - Cataphora's analytics could have automatically processed the relationships between all of the trees, hills, valleys, gullies, poisonous snakes, rivers, and even weather to provide Dante with a map to safety.
Why is C-Analytics important to you?
With so many different means of communication and interaction now possible for many people, more and more "nuggets" of electronic data are being created every day. This leaves a complex trail of small electronic bread crumbs that can provide immense advantage … But only if you are able to see it.
At last, you can make sense of the mass of electronic "nuggets" to ask better questions:
- Is the collection complete?
- Is the other side's production complete? Is there a pattern to what is missing?
- Which topics generated the greatest anxiety and fear?
- Who made the real day-to-day decisions?
- Which processes were consistently followed? Which weren't?
- With whom does a given person share secrets?
- When did a key actor first show signs of disgruntlement?
- What irrelevant, but damaging, content might blindside you?
- What did a particular individual know, and when did they know it?
- Was anyone kept "out of the loop"? Was this deliberate?
And get better answers:
- Verify that you collected data from everyone that you should have
- Identify collections that are not here that should be here
- See channels of communications and know the 'Players'
- Know which files to review first - based on your criteria
- Increase understanding from all the information at hand
What is C-Analytics?
C-Analytics is Cataphora's computer analytics platform to reveal the true nature of facts contained in evidentiary data by calculating combinations - and combinations of combinations - that are quite simply unavailable with other applications. While other products apply simple math and basic calculation, C-Analytics is able to reveal the true nature of previously unrealized fact-patterns by applying the combination of advanced mathematics, technologies, linguistics, and professional investigative methods to combine and parse groups of known facts.
C-Analytics reveals behavior patterns and inter-relationships of people, events, and communications. The results are presented using vivid graphics and easy to understand tables detailing the salient facts of the matter to highlight "big picture" elements that cannot be found using traditional search and review techniques.
C-Analytics does not offer a "smoking gun" or "magic bullet" scenario, but rather it offers insight into the preponderance of the evidence. With analytics, there will be no single document that will prove culpability or a deliberate pattern of behavior - that should remain in the domain of simple search engines and manual review - nor will simple quantitative techniques be adequate. Different types of analyses work together in order to paint a picture of accurately reconstructed reality.
Who currently uses C-Analytics?
Cataphora's assortment of clients includes:
- Corporations - including several Fortune 400 companies
- Law firms
- Financial institutions
- Consultant firms
- Accounting firms
- Investigations firms
Who would benefit from using C-Analytics?
Anyone interested in making sense of volumes of electronic data relating to:
- Civil or criminal litigation
- Discovery compliance
- Government or internal investigations
- Data retention and destruction policies
- Regulatory compliance
How does C-Analytics work?
It's not just the content - you want to look at communication patterns. Cataphora lets you see how and when people are connected…that lets you know where to look for the content you want.
- Esther Dyson
C-Analytics is a powerful framework for building analytics through the use of numerous unstructured data sources (the "nuggets" of electronic data), including email, electronic documents such as Word and Excel, instant messaging, and hand-held devices. The C-Analytics framework enables multi-dimensional modeling, measurement, and visualization of almost endless arrays of fact clusters. These build upon Cataphora's library of existing analytics and model new ones for your particular use-cases. Cataphora's rich suite of automated analytics reveals patterns among and between items of all types, including qualitative aspects of communications, such as tone of voice, relative power or influence of individuals, who knew what/when, and who tended to delete particular types of data, including email. It also shows various types of communication and related insights such as "ad hoc workflows" and "circles of trust" that can be invaluable in gathering facts to support a legal case.
C-Analytics can be used in conjunction with Cataphora's C-Evidence review platform or C-Relevance automated review.
Discussions: the basis for C-Analytics
Common linguistic and statistical methods share a fundamental weakness: they ignore the context of messages. We send emails, make phone calls, and request meetings in response to events, and other people respond to our communications. The meaning of a message can be lost when it is viewed by itself. As a simple example, consider an email that says only "Let's do it!" What is the "it" that is intended here? Let's have lunch? Or let's commit a crime? Certainly the sender of this message intended something specific that the receiver would understand, but without knowing the context, we cannot know what that message was.
This is where discussions come in. A discussion is a conversation about some topic. It has a duration, it includes multiple participants (also called actors), and it may include multiple channels such as email, documents, voicemail, blogs, instant messages, etc. Cataphora's patented C-Analytics software takes these disparate items and reconstructs the discussions that they comprise. These discussions reveal layers of meaning in individual messages, but more importantly, they open the door to understanding a rich variety of patterns within the data. These patterns tell us about day-to-day events. These patterns--and more importantly, deviations from these patterns--can answer questions applicable to a wide variety of legal and investigative scenarios.
Examples
Here are some of the domains included in C-Analytics, and examples of each.
Communication analysis
The pattern of messages moving through an organization reveals a great deal of information not discernable from the content of those messages alone. The activities of individuals and groups, the roles they play within the organization, and the way activities are conducted can be deduced by examining communication traffic. Events such as the coming and going of actors, changes in business practices, and changes in attitude (such as increased caution) all can cause deviations from well established patterns.
- Command structure analysis: The organizational chart says that Jim should have final say on all logistical decisions. Is this the case? How hands-on/hands-off is Jim as a manager?
Formal hierarchies just do not give the whole picture. Who makes decisions depends on managerial style, employee initiative, and on intentional bypassing of the chain-of-command. A clearer picture emerges when you analyze the electronic discussions of the members of a group. C-Analytics detects who typically issues detailed orders, and who just passes on requests.
- "Call me" events: A savvy employee knows not to discuss bad behavior in an email. Is there anything that C-Analytics can still detect?
Often the most interesting-or incriminating-events are the things not committed to written form. When a conversation goes "offline," various clues may be left behind. Occasionally there is direct evidence such as an instant message saying "We need to discuss this in person." Often this is not the case. Extensive and sophisticated automated analysis of patterns in discussions can identify the "gaps" where such events have occurred. Even when the content of a meeting conducted by phone or in person cannot be recovered, knowing about it may help in deposing witnesses.
Proximity analysis
These are techniques that indicate the strength of the professional and/or personal relationship between two actors. These relationships affect investigations in various ways. A relationship may in itself be a fact to verify ("I don't interact much with Jim"), or it may determine another person worth investigating.
- Actor proximity: You know that an illegal activity has occurred, and you've identified at least one individual involved. Where do you go from there? The people that individual works with most closely may be worth investigating as well, but who are they?
Actor proximity is a quantitative measurement of the social and professional interaction between actors. It includes not only frequency of communication, but a variety of factors deducible from discussion analysis and linguistic tools.
Behavioral analysis
These are techniques which measure the affinity of actors for topics and individuals. These show behavior, attitude, and role within the organization. A change in such things probably indicates a major event: perhaps office politics, or perhaps a budding conspiracy.
- Social network analysis: With whom does John interact when he's making operational decisions? How about when he's discussing HR issues?
With whom does a given actor interact on a regular basis, and in what way? Merely counting emails gives a skewed and inadequate picture. Mailing list spam, and messages intended for the different "hats" an actor wears can distort the picture. By combining discussions with linguistic tools such as ontologies, C-Analytics can tease out the facts.
Workflow analysis
The history of a document often tells you far more than the final product alone. What stages did a document go through from origination to release? More importantly, how did that compare to similar documents? Seeing how documents were edited over time shows changes in emphasis and content introduced by various parties.
- Document life-cycle: The sales representative deliberately circumvented sign-offs from business development and finance. The CEO did not sign off on a particular deal. Is there any way to detect these irregularities?
By combining document version detection with discussions, C-Analytics can detect normal workflow for purchases, proposals, etc. Any individual document of interest may be compared against this baseline. Deviations may indicate suspicious behavior.
Deletion analysis
A time honored way for a guilty party to respond to an investigation is to tamper with the evidence. Electronic documents can be destroyed as well as paper documents can. Is there any way to detect this? In the simplest cases, there may be other copies on other computers. But even when this is not the case, Cataphora's advanced C-Analytics software can detect evidence of tampering.
- Unresolved event detection: Once a discussion has been assembled, links to items that should be there but are not become apparent. Some of these are innocent: some items may exist on media not being examined, for instance. Unexplained gaps may indicate selective deletion.
|
