What is IBM Watson
IBM Watson is a IBM supercomputer which uses cognitive computing and lot of service build with AI like NLP, Speech to Text, Text to Speech, Retrieve and Rank, Conversation etc. Initially It was build for only natural Language processing algorithm .Gradually lot of service added to solve different problem like speech to text Tone analyzer and so it uses supervised, unsupervised and Deep learning algorithm as it fit into the model.
Hardware used to build IBM Watson
The system is workload-optimized, integrating massively parallel POWER7 processors. Watson employs a cluster of ninety IBM Power 750 servers, each of which uses a 3.5 GHz POWER7 eight-core processor, with four threads per core. In total, the system has 2,880 POWER7 processor threads and 16 terabytes of RAM.
Waton can process 500 gigabytes, the equivalent of a million books, per second.
IBM apply cognitive learning:
IBM apply different machine learning algorithm to train data and apply cognitive computing on top of it. As a result the answer received for any query from IBM Watson improves and provide better result day by day.The better result of Cognitive computing depends on how we train our data initially like the same way a child Who is mould better under 6 years act smartly as he/she grows over time.
What is cognitive computing:
Cognitive computing is the simulation of human thought processes in a computer. Cognitive computers use machine learning algorithms to continually acquire knowledge from disparate data sources and then present the information as actionable material.Cognitive computing is a very old technique and is getting momentum only recently with technological developments in data mining, pattern recognition and natural language processing.
How to use IBM Watson.
Do we need to buy IBM Watson Supercomputer to build application for our business using its core services? The answer is no.
We use IBM Watson using credentials provided free of cost for 28 days for beginners with Limited resources. We can choose different package as per our requirement to work further. Look into details of pricing at below mentioned link:
IBM Watson Analytics for everyone:
If we want sales team or marketing team work on data and get insight of hidden pattern,We do not need to be data scientist to do that. Instead we can collecte CSV data From database or get social media data like facebook data or twitter data from Watson analytics Dashboard and discover pattern required within few minutes.
For that we need to understand Data ,discovery and display from IBM Watson analytics dashboard.
Go to the link https://www.ibm.com/watson-analytics and create one id and login to understand Ibm Watson analytics.
IBM Watson for application Development:
When considering IBM Watson for developer you can choose services available to build application.
Below are some services widely used by startup and industry for application development and automation .
a)Watson Conversation :
Chatbots are everywhere now a days for getting marketing leads, interacting with customer for getting customer inputs and reply to Query as if human being in doing the conversation. IBM Watson service “Watson Conversation” can be use to create a chatbot in 10 minute .When designing chatbot,you need one chat window, common dialog and intent to work with customer and information stored in your backend so that bot can get relevant answer for customer query.
IBM Watson conversation is build with ready made interface to customer ,how to train the chatbot with dialog and intent and Easily integrate with multiple social and messaging channels including Slack, Facebook and Twilio. .
As we deploy out chatbot ,initially train data is not sufficient and the chatbot could not answer lot of question. Watson recommends what to train next based on frequently asked but unanswered questions.
Login to below link to understand how to build a chatbot for begineers.
b)Watson Discovery :
With IBm Watson we do discovery by uploading data collected as csv format or social media data and discover the insight with
Different pattern (is possible with automatic graph plotting of data as bar diagram,pi-chart and many more).
When we work with R or python program as data scientist ,we do collect data ,understand the data ,understand the problem
Domain,edit missing data and apply mahine learning algorithm to find the hidden insight.
But with Watson we need to only understand the problem and data and upload it for discovery where we can choose the kind of pattern we want to get out of huge data in minutes with various graph plot and share with shareholders or key team.
Please follow the link below to look into more details of discovery.
With deep learning ,it is possible to identify image .For example we can apply neural network to identify whether
The animal picture is a cat or dog.
For example of deep learning ,I suggest going through the following blog:
When we use IBM Watson ,we simply can use Watson API as service visual recognition to identify image instead of writing deep learning code.
Please follow the link mentioned below to see how IBM Watson allow visual recognition .
Speech to Text ,Text to Speech
d)Similarly there are service like speech to text and text to speech to understand speech and convert to text which is more than 94 percent accurate.We can apply these speech services to create customer support virtual agent Chatbot.
Again there is service like personality Insights to predict personality characteristics through text and Tone analyzer to understand emotions and communication style in text .
There are other services available with IBM Watson as you can see in the screenshot.
The demand of IBM Watson :
IBM Watson is widely used in industry to solve problem and lot of startup has come up with analytics solution for business need across various domain.
The magazine https://www.computerworlduk.com explains 18 companies using IBM Watson innovative ways to solve problem in better way.
Competitor of IBM Watson :
IBM is facing tough competition from some startup who are also developing product based on cognitive computing and backed by
gaint like google .Some of them have been acquired by IBM but others are also investing huge in RND to provide better result in
Healthcare like cancer treatment.
Check the link to read details of those competitor and what product they are building:
Do you think AI or IBM Watson can replace human being:
The answer is absolutely no. The AI will reduce the effort of human being .The work those required little effort by hman are automated with learning algorithm. So human will do superior job in 21 century. just like in 90’s bank started working on computerization ,20th century started with complete bank automation with centralized server ,so now the revolution is to replace low quality job with A.I .However the progress is very slow as human being need to develop algorithm for every task they want computer to do. Its not easy.The computer science took 50 years to identify picture with deep learning with accurate algorithm.
Over time Our thinking horizon will change . Kevin Kelly has written a nice article on ““Through AI, we’re going to invent many new types of thinking.”
About the Author
Author of this article “Jamie Rahman” has worked with IT industry over 12 years experience and spent many years
with IBM server for product development.
If you are not comfortable learning be reading and practising yourself ,you can take a course
on IBM Watson analytics at you convenient time and pace by applying the course with below link.