The idea for Info Assistants has been brewing for a while. We started with InfoMinder – a tool for tracking web pages, more than a decade ago. It is still going strong. We addedTopicMinder, a couple of years laeter, to gather news about specific topics using RSS feeds.
Along with tools for gathering information, several technologies for analyzing information are becoming mainstream. The most exciting areas of development include Machine Learning, Deep Learning and Reinforcement Learning. They enable an entirely new generation of Information Assistants.
As Information explodes we need new ways to consume this information. If there are 100 rows in a spreadsheet, you can just browse the information. If there are tens of thousands of rows of data, you can slice, dice, build pivot tables and apply aggregate functions. What do you when you have millions of points of information. ML/DL techniques come to the rescue.
How do you make complex information easy to consume? Some visualization techniques will help. Finding the right visual abstractions are not, however, easy. An emerging area for reducing complex interactions with information is a conversational interfaces. Thanks to rapid improvements in natural language understanding and generation, we now have very powerful conversational interfaces. They are in their infancy, but hold a lot of promise. NLU and NLG are both powered by ML and DL techniques.
So we are recasting some of our old tools into new groups. Information gatherers (infotools), information analyzers and infobots. The tools help you discover and gather information. Analyzers help you make sense of information. And Infobots provide a friendly chat powered, conversational interfaces to information.
