Google has found to be awarded different patents on interesting topics like machine learning, knowledge panels and so on. Knowing these patents would only result in an expansion of knowledge about the digital world. If the aim is to improve the hold over SEO or Search Engine Optimisation, then knowing these patents could be really worth.

In this write-up, we shall discuss a few Google Search related patents that are mentioned in the Search Engine Journal. Given below features some details on these patents:

.

Machine learning to identify opinions in documents

This patent was filed in April 2020 and was awarded recently on the 10th of November. This patent is about the systems that employ a machine learning opinion classification model that classifies the portions of the documents. The documents stated here could be news articles web pages, etc., whereas the clarified portions are sentences, phrases, paragraphs, etc. The various portions of the document can be selected as inclusion for information display depending upon their importance and ranking. This patent, just like the other useful patents, can be used in various ways; however, the major focus falls on the news items or potential blogs. As per one statement in this patent, a particular portion of the document can be classified as an opinion when it is being ranked higher due to its high importance.

According to this patent, as only the important information would reflect on the display, there are chances that the exposure to new information could get limited. As per this patent, it would be really difficult to understand the content in the document by using a computing system. If talked about the content featured in high profile articles written by professional journalists, the content here would be in a high-quality language, and there are possibilities that the computing system would be able to understand just a little out of it.

The task of determining or comparing the articles of different journalist would become even more difficult.

Terms, Category and Link-based User Profiles can be used for Personalization of Web Search Result

This patent was filed in March 2020 and was rewarded in December. If we take a deeper dig about understanding this patent on personalization, it basically features PageRank or Random Suffer approach, including its limitations too. Its emphasis on shifting towards the more personalized approach and an improved profile rank to attain superiority of the existing search result. A query or an opinion submitted to the search engine is based on the preference of the user. The quality of the result that the user receives depends upon the satisfaction of the user about his or her query. It is found that when the user's query is clearer or better defined, then it is more likely that he or she would get better quality or satisfactory results. However, if the user's preference is not properly defined, looks more like a random surf, or seems biased due to some personal reasons then the search results obtained from the same search engine would seem a bit less useful or less knowledgeable. This implies that if the user's query feed to the search engine should be precise but not completely different from the random surfer's query if the aim is to achieve the most satisfactory of most useful search results.

Application Environment consisting of Contextually Relevant Knowledge Panels

The filing date of this panel is May 23, 2019 and it was awarded on the 15thof December in 2020. This patent features almost all the new elements that are familiar to most of us when it comes to discussing the knowledge panels. However, there are hardly times when we come across such descriptive patents, and this is the reason why it is being featured in Search Engine Journal too.One thing that this patent makes us learn is when it comes to describing something which does not exactly know them, it can be stated as an entity. Entities are a broad phenomenon and are not limited to just including places, people, animals, landmark, historical events, country, businesses, sports, movie, song, album, sports teams, sports event, an art form, game, or any other thing besides this.

As per one abstract of this patent, the content given in the document contains the factual entity which relates to the actuality of it. Requesting content from the knowledge panel could prove to be useful here. This is because the knowledge panel is nothing but the kind assemblage that contains a collection of content that perfectly relates to the factual entity the user has been looking for. Thus, the knowledge panel can be regarded as the user interface element. The content requested on the knowledge panel is contemporaneously displayed on the user device along with the content of the document.
The patent further states that, the user would often need to provide the contextual information of the document in the particular query when it comes to developing search queries for submitting to the search engine. If described in the simplest way, the formulation of the query from the user's end should be in such a way that it expresses the informational need.

There are also some implementations according to which the knowledge panel provides a proper summarization of the information relating to a particular entity. For example, if the knowledge panel is considered for a singer, then it may include details such as an image of the singer, description of the singer, some interesting facts about him or her and the content that features his or her songs or albums.

One thing that we must acknowledge is it is always better to read more and more about the patents, trends or any other information related to Google. The reason behind is, it could act as a potential source to stay prominent on the digital platforms. The more we read, the more we would get to know about it, and the more it would get easier for us to rule.