Saturday, March 25, 2017

IBM Watson Cognitive Technology‎ - How it Works?

Introduction


Today we are heading to the smarter computer technology, many IT companies are trying to keep ahead of the trend and figure it out what is going to come in the near future. among all of the emerging technologies, the Artificial Intelligence seems to capture the pulse of the new technology, most businesses realized how important it is to embrace the AI human brain power to compete in the 21st century. They realized the coming of AI could be the new Electricity that powering the future economy and penetrating every aspect of our life.

Introducing the breakthrough technology that IBM has been developed more than a decade ago called "IBM Watson Cognitive Technology", unlike the conventional computing approach based on rules and logics, it is a new type of computing called "Cognitive Computing’, just like how our human learn new things and draw a decision. Watson is pretty mature and powerful technology at this moment. If it uses carefully by the human linguist experts, it will unleash the potential human brain by keeping learning the new things of the feeding training data and countless self-iteration process.

You will find it useful in the medical field, law, Question Answers and even the cooking recipe domain. If you master the tool and its on-going machine learning algorithms, I am sure your business will be on the right path with many success for the years to come.

How does IBM Watson work?


Watson works just like how human study the new information and make a decision based on the logics of the data. First, let"s understand how our human study and decision work?

1) We observe the new information and trying to understand the rules, logics behind it.

2) We interpret it and make variable Hypotheses based by it

3) We evaluate the Hypotheses to decide which is right or wrong

4) We make the final decision based on the evaluation and act upon it


IBM Watson works as a similar approach:


1) It reads unstructured data in a massive speed and scales just like a human being reads on a daily basis, it breaks down the sentence dramatically, relationally and structurally to discern the semantic, idiom, synonym meaning of the written materials.

2) With a guidance by a human expert, Watson collects the required knowledge of literacy in the specific domain, this is called "Corpus". Building the smart Corpus also need the on-going human intervention to sort out the good or bad information into the system, the process is called "Curating the Content".

3) Next, the data is preprocessing by Watson, building the efficient metadata within a specific area of content, the process is known as Digestion. Watson is also creating a knowledge pattern to answer questions accordingly.
4) Now the Corpus metadata is ready, it needs to partner with a human language expert to upload the data in the form of question & answers pairs which trained by "Machine Learning" Algorithms. This process now doesn"t give Watson the smart answers but rather teaches the meaning of the linguist pattern in this domain. Watson will continue to learn and update the new information that is available in the cloud by the intervention of the human expert hands. Now finally Watson is ready to answer the questions based on its intelligence, evidence and the massive scale of big data.
5) Among all of the Watson application approaches, there are the common methods that translate how Watson in action?

- After it understands a specific human sentence, it generates the Hypotheses.

- It then looks for the evidence to support or refuse the Hypotheses

- It scores each hypothesis based on the statistics modeling patterns known as "weighted evidence score"

- It estimates how much probabilities the evidence is during the evaluating session

- Finally, a human expert reads the data to make a better decision from its analytics inspiration


IBM Watson Cognitive Technology‎ Features:


- It can understand all form of data and pattern on a large scale

- Unlike the conventional computing, it is a new type of cognitive computing that modeled how human learned the new things.

- It understands the unstructured text, images, audio, and video.

- It understands the human reasoning, personality, tone, and emotion.

- It keep learning new information with the machine learning algorithms by the minor human intervention

- It can interact with your users by the build in linguist model and speech recognition technology

- Watson has built in API and set of its SaaS products for businesses to unleash their brain potential

- It has well documented boilerplate, sample code for developers to get started it in no time.

- Its starter kits is great for developers to build some cool cognitive apps

- It is affordable and cost effective for small businesses

IBM Watson API


If you are a WebApp developer, you can capitalize the ready used Watson API to help you get started it and build some cool artificial intelligence Apps.

The API included the Alchemy Language, Conversation, Dialog, Document Conversion, Language Translator, Natural Language Understanding, Personality Insights, Tone Analyzer and so on.

If you don"t want to start everything from a scratch, API is a great easier way to build you a first smart and cognitive app by leveraging the IBM Watson cloud data, it supports programming languages like NodeJs, Java, Python etc.

Learn More At: https://www.ibm.com/watson/developercloud/services-catalog.html


IBM Watson areas of uses:


IBM Watson debut in Jeopardy Show

The practice match was on January 13, 2011, Watson won a 15 questions round with the score of $4,400 that slightly performed better than its two competitors named Jennings and Rutter.

The first round of match was on February 14, 2011, Watson, this time represented by a computer monitor with the artificial voice was a tie with Rutter at $5,000. After the first round, Watson placed second after Rutter and Jennings.

The Second round of final outcome - As a result, Watson won the game with a score of $77,147 to receive the first place 1 million rewards which IBM has promised to donate the money to the charities.

IBM Watson Cognitive Technology in Healthcare

Watson can be used to analyze data in your medical organization to drive better performance, study massive scale of your patient’s symptoms based on the science research database to increase the odds of overcoming the disease treatments.

It supports your medical team with smart insights, to combat Hypotheses and draw a better decision-making process.

Watson understands patients better than himself by its on-going data training and iteration learning process. It improves the patient overall health experience and reduces the labor cost.

IBM Watson Price


Watson has many services that ranging from Alchemy Language to Data Analytics, the pricing model varies differently from the services. Most services have the free and premium version or based on API calls. You can check on the overall pricing model as the below chart.

http://fredrikstenbeck.com/what-does-watson-cost-what-is-the-price/

Final Words


Jack Ma once said, "if you want to succeed in the business, learn how to think ahead of the curve, things that will happen 10 to 20 years from now". I believe the AI technology is not even far from now, in fact, it is a ready booming industry to come. So whoever masters the skills of how to use this technology will be ahead of the curve and unleashes his company human brain power to a massive intelligence scale of big data, adapting it to the future of competition and benefit from it.

If you have understood what this article really talk about? You should make every effort to learn by yourself or hire a data special to implement the AI technology in your business so that you can benefit in the future just like the old Electricity boom.

[embed]https://www.youtube.com/watch?v=_Xcmh1LQB9I[/embed]

 

[Tweet "IBM Watson Cognitive Technology‎ - How it Works?"]




 

No comments:

Post a Comment