Google published three patent applications on Google Notebook this week, which describe the fundamentals of how the program works, and provide a hint at how notebooks may influence some search results.
The nice thing about the Google notebook is that it has the potential to be a helpful research tool, enabling you to quickly save and organize information that you find on the Web. Having said that, I’ve had it installed on my desktop for many months and rarely find myself using it.
Some of the newest features that aren’t covered in the patent filings include the ability to turn your notes into Google Documents, the mobile version of Notebooks, an integration with Google Maps and with the personalized home page, and the ability to add labels to notebooks.
If you’re interested in some of the finer details of how notebooks work, and the assumptions behind their creation, then you may find some value in looking over the patent applications that describe them:
The third document on “Presenting Search Result Information” hints that rankings of pages in search results may potentially be positively influenced by the inclusion of a link to that page in a Google Notebook. It’s difficult to tell if Google is using Notebook to influence rankings of results.
Furthermore, the higher ranking can be based on an analysis of the contents of the web notebooks 118. For example, a web site matching a title, heading, user-annotation, metadata or clipped content from a web notebook 118 can receive a higher ranking than one that matched no web notebooks.
It’s not just the inclusion of a site in a Google Notebook that would potentially affect rankings, but also a number of factors that may be necessary for such a positive influence to exist.
The web notebook-based ranker 404 uses the content of web notebooks to determine ranking. Web notebook content can include titles or headings in the web notebooks, snippets that have been clipped into the web notebooks, user-supplied annotations or user-supplied free-form text, metadata associated with web notebooks (e.g., metadata that identifies a corresponding search request, a time/date stamp, or other information related to a snippet, annotation, heading, title, etc.), and other information stored in the web notebooks.
How might those factors be weighted? It may depend upon how good a match the notebook entry is for the query used by a searcher:
Ranking can be based on the extent to which the search results match content in existing web notebooks.
For example, the web notebook-based ranker 404 can assign a higher rank to search results corresponding to websites matching content of web notebooks. In particular, if the user enters a search query for “Hawaii vacation” in the web browser 104, the web notebook-based ranker 404 can rank the corresponding search results based on whether existing web notebooks that contain content clipped from the search results have titles matching the keywords (e.g., “Hawaii” and “vacation”) or contain clipped content matching the keywords.
If there are a number of notebooks which share a URL with a set of search results, the quantity of notebooks mentioning the URL, and the frequency of occurance of the keywords searched for which appear within the notebooks might also influence the rankings of that URL in the search results for that query.
In addition, the web notebook-based ranker 404 can assign higher ranks to search results matching a greater number of notebooks and/or a greater number of occurrences of keywords within those notebooks.
In each instance, respectively, the system may presume that notebook authors select descriptive titles for content they clip, that clipped content matching keywords is a more important part of a web page than is unclipped content, or that the number of clippings from a web page reflects its perceived usefulness to users in general.