How Big Data drives Facial recognition accuracy, Social graph performance and Work force productivity

Facebooktwitterpinterestlinkedinmail

Here are the latest developments from the big data world on facial recognition, social graph and company productivity.

Facial signature with big data: Voice of Big data, a US based big data analytics firm is all set to launch a facial recognition product. Facial recognition with the combination of big data analytics is an upcoming technology to identify criminals and help solve crimes. This technology is very similar to some of the latest Hollywood flicks where the computer software rapidly crunches massive amount of data from the database to match facial characteristics of the suspect.

Existing facial recognition software rely on five to six key facial features such as:

  • Nose structure
  • Eye socket positioning
  • Jaw line prominence
  • Forehead and distance between the eyes
  • Shape of the cheekbones

Big data analysis will give boost the accuracy and performance of the match. The advantage of big data hypotheses is relying on N=all sample size. What it means is instead of using five to six comparison points recognition software will be able to use all facial characteristics. This process improves the accuracy of the match quite significantly.

Apache Hive vs Apache Giraph: A big rule of any big data technology is it must carry a name of an animal, i.e. Hortonworks, Hadoop and now Giraph :). Jokes apart, Facebook continues to expand its footprint in compute and big data analytics world. Facebook has chosen Apache Giraph over Apache Hive to power a new type of search based on social graph. Facebook can rely on Giraph to scale up to a trillion edges.

Apache Giraph was developed by Yahoo in 2010 based of off Pregel paper published by Google. Giraph is based on the fundamental of distributed computing and iterative graph processing system. There are three critical components of parallel computation and processing – Concurrent computation, communication and barrier synchronization.

This three-part model provides performance improvement for graph based algorithms.

apache giraph

Source: Apache.org

Smart work force smart data: Social media continues to grow rapidly in the personal and professional sector. There are approximately 1.5 billion social users across the globe and the unlocked social value according to McKinsey Global Institute (MGI) is close to $1 trillion. MGI recently conducted a study to trace the growth of social technologies, examine the source of their power, and understand the ways in which social technologies create value.

In this study, MGI reports the evolutions of the behavior of the social media users and how it expands over their personal and professional life. Social is becoming interwoven in all of our activities right from share personal pictures to harnessing collective knowledge using crowd-sourcing. Yet, only 3% of the 4,200 companies interviewed reported achieving substantial benefits from the use of social technologies.

Based on the in-depth analysis of usage, MGI has identified ten ways in which social technologies can generate $900 to $1 trillion in value.

value from social media

 

Companies can rely on social for creating products using crowd-sourcing, forecast and monitor, derive customer insights, marketing communication, social commerce, customer support, increasing collaboration and matching task to talent.

MGI claims social technologies can improve worker productivity by 20-25% by leveraging enterprise networking tools. The value creation is dependent upon the ease of the data capture. In some cases, capturing data and putting it to use is much easier such as embracing Chatter or Yammer instead of emails. However, other areas such as Energy or Pharmaceuticals industry have siloed disintegrated data sources making it much more difficult to generate value.

The use of social technologies also carries risks and could impact employee productivity if employees are engaged in unfruitful “chatting”. Enterprise could overcome these challenges by establishing governance around the usage of the internal communication tools. Other types of risk involve compromising user privacy and information security.

The entire MGI report can be downloaded here for further reading and analysis.

As always, appreciate your comments, feedback and tweets.

Subscribe
Notify of
guest

Time limit is exhausted. Please reload the CAPTCHA.

0 Comments
Inline Feedbacks
View all comments