With the launch of Google Hummingbird update and the increased use of semantics and Knowledge Graph, the role of entities along with their relationship are set to play a greater role in the ranking of search results. Majority of search queries contains an entity of some sort or the other. For example, a query like "Neil Armstrong Biography" contains an entity "Neil Armstrong". With the help of Knowledge Graph database, Google can easily identify that Neil Armstrong was an astronaut and it can serve the query of the user in a well defined manner. This is the power of Hummingbird backed with the cumulative effectiveness of semantics and Knowledge Graph.
Apart from just identifying entities, there are certain terms related to an entity which provide substantial value to Google in order to process user queries in a more defined manner. As per the example above, the terms associated with the entity might include:
Ohio (where he was born)
United States (his native country)
Gemini 8 and Apollo 11 (spacemissions)
Carol Held Knight and Janet Shearon (his wives)
Eric Armstrong, Marc Armstrong and Karen Armstrong (his children)
Nasa, aerospace engineer, first person to walk on moon, astronaut (terms used to associate Armstrong with)
These terms are great signals for Google to identify the closest matching entity.
Now, Google can Think Both Ways
Search terms associated with an entity and the relationship between them has empowered Google to think both ways! Now, if a search query contains an entity then Google can easily identify information to that entity and if a search query does not contains any entity then still Google can identify entities based on the processing of the terms associated with those entities and those contained in the search query. A recent patent granted to Google has provided enough vocabulary to Google in order to sharpen up its brains using its popular Hummingbird algorithm along with a perfect match of semantics.
The patent granted to Google on November 19th, 2013 discusses about "Assigning terms of interest to an entity". The patent lets Google identify candidate terms and weigh each one of them as per their relative frequency.
Use of Data Mining
Under data mining, patterns of data are identified and processed in order to enhance intelligence and determine true value of concepts.
Steps to Process Queries
1- Identification of candidate terms - As a first process, some candidate terms are identified using filtering, modifying and scoring techniques.
2- Identifying known for terms- Now, an evaluation of candidate terms reveals terms that the entity is known or recognized for. It may describe features, unique features associated with the identity, location of the entity, understand more about the entity etc.
3- Continuing the process of identifying semantics of known for terms associated with the entities.
4- Identification of known for terms in different languages.
Here is a detailed description to the patent:
Full Patent Information Can be Viewed Here:
Assigning Terms of Interest to an Entity
Also See:
How Google Identifies Substitute Terms of a Query?
Google Patent to Identify Erroneous Business Listings
How Google Identifies Spam in Information Collected From a Source?
Google Patent Named Ranking Documents to Penalize Spammers
Taxonomic Classification While Finding Context of Search Query
Google Granted Patent for Detecting Hidden Texts and Hidden Links
Rich Snippets in Google
How to Add Ratings and Review Stars on Google Search Results
Query Highlighting on Google Search Results
List of Google Search Operators
How Does Google Identifies Terms Related to an Entity?
Apart from just identifying entities, there are certain terms related to an entity which provide substantial value to Google in order to process user queries in a more defined manner. As per the example above, the terms associated with the entity might include:
Ohio (where he was born)
United States (his native country)
Gemini 8 and Apollo 11 (spacemissions)
Carol Held Knight and Janet Shearon (his wives)
Eric Armstrong, Marc Armstrong and Karen Armstrong (his children)
Nasa, aerospace engineer, first person to walk on moon, astronaut (terms used to associate Armstrong with)
These terms are great signals for Google to identify the closest matching entity.
Now, Google can Think Both Ways
Search terms associated with an entity and the relationship between them has empowered Google to think both ways! Now, if a search query contains an entity then Google can easily identify information to that entity and if a search query does not contains any entity then still Google can identify entities based on the processing of the terms associated with those entities and those contained in the search query. A recent patent granted to Google has provided enough vocabulary to Google in order to sharpen up its brains using its popular Hummingbird algorithm along with a perfect match of semantics.
The Patent Empowering Semantics, Hummingbird and Knowledge Graph
Use of Data Mining
Under data mining, patterns of data are identified and processed in order to enhance intelligence and determine true value of concepts.
Steps to Process Queries
1- Identification of candidate terms - As a first process, some candidate terms are identified using filtering, modifying and scoring techniques.
2- Identifying known for terms- Now, an evaluation of candidate terms reveals terms that the entity is known or recognized for. It may describe features, unique features associated with the identity, location of the entity, understand more about the entity etc.
3- Continuing the process of identifying semantics of known for terms associated with the entities.
4- Identification of known for terms in different languages.
Here is a detailed description to the patent:
Full Patent Information Can be Viewed Here:
Assigning Terms of Interest to an Entity
| Inventors: | Lee; Jason (Forest Hills, NY), Stern; Tamara I. (New York, NY), Donaker; Gregory J. (Brooklyn, NY), Blair-Goldensohn; Sasha J. (New York, NY) | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Applicant: |
| ||||||||||
| Assignee: | Google Inc. (Mountain View, CA) | ||||||||||
| Appl. No.: | 13/430,624 | ||||||||||
| Filed: | March 26, 2012 |
Also See:
How Google Identifies Substitute Terms of a Query?
Google Patent to Identify Erroneous Business Listings
How Google Identifies Spam in Information Collected From a Source?
Google Patent Named Ranking Documents to Penalize Spammers
Taxonomic Classification While Finding Context of Search Query
Google Granted Patent for Detecting Hidden Texts and Hidden Links
Rich Snippets in Google
How to Add Ratings and Review Stars on Google Search Results
Query Highlighting on Google Search Results
List of Google Search Operators
No comments:
Post a Comment