Searching Social Graphs To Clarify Conversations Across The Web (And Spell Check)


Search engines and technology companies building ad targeting platforms will pay attention to social connections and words used on the Web far more than paid search ads or recommendations.

A recent update for a patent application from Google describes how the search engine might add words to a dictionary on a PC or a mobile device from a search query through textually disambiguation using social connections. The theory is that one person might likely use terms their friends often use. From this information, and more, collected across the Web, search engines, in theory, could build a social graph that maps likes and dislikes of a specific person or persons, as well as those whom they friend.

These words collected that create "textual disambiguation using social connections" come from social networks where the person might be a member, describes SEO by the Sea founder Bill Slawski. Describing the patent application to demonstrate the power of social graphs, he highlights the mounds of information search engines capture about someone's use of words on the Web, as well as their social connections.



The patent filing details how Google might score words used by people in a social network to decide on which ones to add to a dictionary. This score aims at predicting future words someone might use. Slawski tells us the method for deciding on words to present as suggestions could also look at information found from someone's computer, taken from word processing documents, calendar items, contacts, history from a browser, and more.

Slawski describes it this way: If someone frequently visits the baseball pages at ESPN and those files are in their browser's cache, when they start spelling b-a-s, the computer they are using might offer "baseball" as a query suggestion.

This Google patent seems to focus more on the auto-completion of words on a mobile phone than it does on providing query suggestions to a searcher. Slawski says the information located on a network that someone uses, such as their email account, might also become a source of data that could help someone fill out a text box on their phone, or in suggesting a query term. But could it also help advertisers target consumers and their friends in a social graph.

Although not a new concept, the technology has matured and ready to deploy across the Web as an alternative to behavioral targeting. Ex-Googler and Media6Degrees chief executive officer Tom Phillips calls it social targeting. And he's out to prove it works. The company has developed technology-social CRM-that will dynamically serve up display ads based on your social graphs. The non-personally identifiable (PII) data collected in browsers link to others and make a decision on the content to display within milliseconds. When I asked Phillips when we can expect to see the technology, he told me his company's board has been asking the same thing.

As I noted, the concept isn't new, but rather the technology that supports a link between search and social graph connections has gained enough maturity to work. HuoMah SEO Blog founder David Harry points to a Microsoft patent filed in September 2006.

Next story loading loading..