How Google May Annotate Images to Improve Searches

Google might improve on information from sources like knowledge bases to them answer search queries. The information may be learned from sources outside of those knowledge bases when Google may analyze and annotate images and consider other data sources. A Google patent that came out recently on this topic defines knowledge bases for us and why those are important.

Analyzing Images to Enrich Knowledge Base Info

This approach may annotate images and select entities contained in those images. This is an effort to understand better as well as interpret images and explore related entities in images, so Google can focus on the relationships between the object entities and attribute entities, and store those relationships in a knowledge base. Google can learn based off of pictures of real-world objects. When writing about mapping image queries, you may think about labels helping to organize information in a useful way. Use those labels to learn about entities when doing content creation or keyword research. Image searching and looking at those semantic labels can be worth the effort. This new patent lets us know how Google might assign annotations to images to identify entities in the images. While labeling, they may choose an object entity from the entities available and then select at least one attribute entity from an annotated image that has the object entity. They could also conclude a relationship between the object entity and attribute entity or entities and include that relationship in a knowledge base.

Defining an Attribute Entity in an Image

An attribute entity may be one that is among the entities contained in images that contain the object entity. They’re entities other than the object entity. Annotated images that have the object entity may be grouped, and an attribute entity may be chosen based on what entity may be contained in the most significant number of grouped images as identified by annotations. A group of 30 annotated images containing the object entity “George Washington” may also have 20 images that have “Martha Washington.” In this case, “Martha Washington” may be considered an attribute entity.

Inferring Relationships between entities by Analyzing Images

If any images of Michael Jordan has a basketball in his hand, a relationship between Michael Jordan and basketball might be made (That Michael Jordan is a basketball player.) By analyzing those images of Michael Jordan with a basketball in his hand and wearing a jersey for the Chicago Bulls, a search query asking a question such as “What basketball team does Michael Jordan play for?” may be completed with the answer “Chicago Bulls.”

To answer a search query such as “What team did Michael Jordan play basketball for,” Google could search an image for “Michael Jordan playing basketball.” Having images that contain the object entity of interest can let the images be analyzed, and an answer will be provided.

Contact BWM

Boston Web Marketing can help you, and your business stay up-to-date on the latest SEO trends! Give us a call at 857-526-0096 or fill out our online contact form. We look forward to optimizing your sitesocial media and help you get found quick!

Recent Blog Posts

Contact Us Today!

  • This field is for validation purposes and should be left unchanged.