Exploring how Google’s knowledge graph works can provide some insights into how is growing and improving and may influence what we see on the web. A newly granted Google patent from the end of last month tells us about one way that Google is using to improve the amount of data that its knowledge graph contains.
The process involved in that patent doesn’t work quite the same way as the patent I wrote about in the post How the Google Knowledge Graph Updates Itself by Answering Questions but taken together, they tell us about how the knowledge graph is growing and improving. But part of the process involves the entity extraction that I wrote about in Google Shows Us How It Uses Entity Extractions for Knowledge Graphs.
This patent tells us that information that may make its way into Google’s knowledge graph isn’t limited to content on the Web, but can also may “originate from another document corpus, such as internal documents not available over the Internet or another private corpus, from a library, from books, from a corpus of scientific data, or from some other large corpus.”
What Knowledge Graph Reconciliation is?
Continue reading “Google Knowledge Graph Reconciliation”
“How Long is Harry Potter?” is asked in a diagram from a Google Patent. The answer to this vague question is unlikely to do with a length related to the fictional character but may have something to do with one of the best selling books or movies featuring Harry Potter.
When questions are asked as queries at Google, sometimes they aren’t asked clearly, with enough preciseness to make an answer easy to provide. How does Google Answer vague questions?
Question answering seems to be a common topic in Google Patents recently. I wrote about one not long ago in the post, How Google May Handle Question Answering when Facts are Missing
So this post is also on question answering but involves issues involving the questions rather than the answers. And particularly vague questions.
Continue reading “How would Google Answer Vague Questions in Queries?”
I wrote about a similar patent in the post, Google Extracts Facts from the Web to Provide Fact Answers
This one introduces itself with the following statement, indicating a problem that Google may have with answering questions from the facts it may collect from the Web to fill its knowledge graph:
Embodiments relate to relational models of knowledge, such as a graph-based data store, can be used to provide answers to search queries. Such models describe real-world entities (people, places, things) as facts in the form of graph nodes and edges between the nodes. While such graphs may represent a significant amount of facts, even the largest graphs may be missing tens of millions of facts or may have incorrect facts. For example, relationships, edges or other attributes between two or more nodes can often be missing.
That is the problem that this new patent is intended to solve. The patent was filed in November of 2017. The earlier patent I linked to above was granted in June 2017. It does not anticipate missing or incorrect facts like this newer patent warns us about. The newer patent tells us about how they might be able to answer some questions without access to some facts.
It’s also reminding me of another patent that I recently wrote about on the Go Fish Digital Website. That post is titled, Question Answering Explaining Estimates of Missing Facts. Both the patent that post was about and this new patent include Gal Chechik, Yaniv Leviathan, Yoav Tzur, Eyal Segalis, as inventors (the other patent has a couple of additional inventors as well.)
Continue reading “How Google May Handle Question Answering when Facts are Missing”
Google Has Filed A Patent for Location Extensions to Geographical Paid Ads
When a Google Patent uses the word “Content” they often mean advertisements, rather than just the content on a website. That was true when I wrote about a Google patent about combining advertisements and organic results in the post Google to Offer Combined Content (Paid and Organic) Search Results
I don’t often write about paid search here, but sometimes see something interesting enough to write about. We’ve been seeing some of the features from organic search appearing in paid search results, such as sitelink extensions, and Structured Snippets extensions. Google has written up extensions, which are ways of adding additional information to advertisements “to maximize the performance of text ads.”
One specific type of extension is are location extensions. Location Extensions can add information to an advertisement that you bid upon that can exhibit more information to your ad, such as:
Continue reading “Location Extensions Augmented Advertisements”
A New Patent on Categorical Quality
Some of the people who write patents for Google tend to stand out to me. One of those is Trystan Upstill. I noticed that he has published another one that looks really interesting, and worth reading. When I started following his patents, I read his doctoral thesis, Document ranking using web evidence which was really interesting, from the early days in his professional career. It is from before he was listed as the inventor of a number of patents, that I also found interesting. I’ve written about a number of patents he has participated in creating as well because they often focus upon Site Quality, and I learn something from reading them and trying to understand them. Here are posts from his patents which I have written about previously:
I noticed his name on a new one granted at the end of May, and I’ve been working through it now, too.
Continue reading “How Google May Rank Some Results based on Categorical Quality”
Image Classification in the past
Back in 2008, I was writing about how a search engine might learn from photo databases like Flickr, and how people label images there in a post I wrote called, Community Tagging and Ranking in Images of Landmarks
In another post that covers the Flickr image classification Landmark work, Faces and Landmarks: Two Steps Towards Smarter Image Searches, I mentioned part of what the Yahoo study uncovered:
Using automatically generated location data, and software that can cluster together similar images to learn about images again goes beyond just looking at the words associated with pictures to learn what they are about.
Continue reading “Image Classification and Landmarks at Google”