Imagine you could copy any part from a web page in your browser and paste it into a notebook in the cloud. Styles and images would be preserved, so that the clipped part looks just the same as on the original page, but it would be kept as you found it, for as long as you want, annotated with the time and place where it was originally found. You could for example keep track of results from booking a trip across multiple web sites on a single notebook, adding your own comments as required, to remember what you did.
In a previous post I asked “How to bring ML to Search?” In this post I want to discuss which companies could likely bring machine learning to search engines. Some ML techniques are already used for Web search, but search engines fall behind modern possibilities by far. When I speak of ML in this post, I also mean related NLP techniques. The benefits of bringing ML to search are many: better spam fighting, a replacement for PageRank, higher quality and more varied results on the first page and search term disambiguation, among others.
There is a Natural Language Processing (NLP) technique called Question Answering (QA), where users ask a question in natural language to a system, and receive an answer. For example we might ask, “what is the maximum take-off weight of the Airbus A380?” and receive the answer “575,000 kg”. This is an improvement over plain search engines, which only take keywords as a query and answer with entire pages or documents of information, from which the user has to extract the desired answer.
I have written about Search Agents before (1, 2). Make sure to read those posts to know what I’m talking about here. This post delves a bit deeper into business models for Search Agents. The key business issue is how to monetize search users with this novel approach to search, while keeping costs from use of third party search services low. So how to make money with a Search Agent? A straightforward way to earn a commission would be to run banner ads or text ads from an ad network on the results pages.
Messaging apps are normally location-independent by design. If your device is connected to the Internet, sending a greeting or a snapshot to any other online device is normal. But what if the spatial dimensions were re-introduced to a chat application? Send a message to anyone listening in a 100 meter radius or to anyone in the same metropolitan area. What if the app guaranteed that you really have to be where you are to send and receive such messages?
If you use leading search engines a lot, you have probably noticed that they haven’t improved much over the years. They really seem to be stuck in the 2000s, just with more spam than back then. One way in which they could improve, is by using Machine Learning (ML) for processing search results. Some of the companies involved have good ML and AI teams, so why can’t they use that know-how to improve their search engines?