SPSKC Machine Learning in SharePoint
Download as PPTX, PDF1 like1,748 views
The document discusses advanced analytics processing techniques for enhancing search functionalities, focusing on aspects like anchor text processing, click distance, search clicks, and social tags. It details how these analyses boost or demote items in the search index based on user interactions and relationships among users. Additionally, it emphasizes machine learning and personalization to improve user experience and recommendations in SharePoint.
SPSKC Machine Learning in SharePoint
- 1. #SPSKC ©2012 Microsoft Corporation. All rights reserved.
- 3. Naomi Moneypenny [email protected] Technology forecasting and strategy manager at Shell, consulted at many Fortune 100 companies since 3000+ followers on Twitter 3+3 dogs 20+ patents in adaptive systems Astrophysicist Passionate about user adoption and enterprise collaboration & innovation @nmoneypenny wwww.Synxi.com geek
- 6. Machine learning is the most significant technology trend. Computers have to get smarter and anticipate. Kevin Turner, Microsoft COO, July 2012 @nmoneypenny wwww.Synxi.com
- 8. ©2012 Microsoft Corporation. All rights reserved.
- 12. Analysis Description Anchor text processing Anchor text processing analyzes how items in the content corpus are interlinked. It also includes the anchor texts associated with the links in the analysis. The Analytics Processing Component uses the results of the analysis to add rank points to the items in the search index. Click Distance The Click Distance analysis calculates the number of clicks between an authoritative page and the items in the search index. An authoritative page can be a top level site, for example http://www.contoso.com, or other pages that are viewed as important. You can define Authorative pages in Central Administration. The Analytics Processing Component uses the results of the analysis to add rank points to the items in the search index. Search Clicks Social Tags The Search Clicks analysis uses information about which items users click in search results to boost or demote items in the search index. The analysis calculates a new ranking of items compared to the base relevance. The clicks data is stored in the Link database. The Social Tags analysis analyses social tags, which are words or phrases that users can apply to content to categorize information in ways that are meaningful to them. In SharePoint Server 2013, social tags are not used for refinement, ranking, or recall by default. However, you can create custom search experiences that use social tags and the information from this analysis. Social Distance The Social Distance analysis calculates the relationship between users who use the Follow person feature. The analysis calculates first and second level Followings: first level Followings first, and then Followings of Following. The information is used to sort People Search results by social distance. Search Reports The Search Reports analysis aggregates data and stores the data in the Analytics reporting database where it's used to generate these search reports: •Number of queries •Top queries •Abandoned queries •No result queries •Query rule usage The report information is saved in the Search service application, and not with the items in the search index. If you delete the Search service application, the report information is also deleted. Deep Links The Deep Links analysis uses information about what people actually click in the search results to calculate what the most important sub-pages on a site are. These pages are displayed in the search results as important shortcuts for the site, and users can access the relevant sub-pages directly from the search results.
- 15. + Follow when you know sources of information that are generally relevant @nmoneypenny wwww.Synxi.com ? Search Discovery when you know what information you need now but don’t know where it is when you don’t know what you need now or even know that it exists
- 18. Class 2 Class 1 @nmoneypenny wwww.Synxi.com Class 2 Class 1
- 22. ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) ( ) (.) Input space Feature space Note: feature space is of higher dimension than the input space in practice @nmoneypenny wwww.Synxi.com
- 24. Recommendations of Content, People, and Topics Relevancy & Quality Recency New to you? Popularity People like you Ratings Inferred Relative Expertise Personalization Topic 1 Topic 2 Topic 3 Topic 4 . . . . . . . . . . . . . Topic N . . . . . . . Topic N Inferred Interests Topic 1 Contextualization @nmoneypenny wwww.Synxi.com Topic 2 Topic 3 Topic 4 . . . . . .
- 25. The Adaptive IT Stack ® Learning Layer Synxi / Personalization Apps Social Layer Social Platforms Process Layer Content & Applications Layer Cloud (Internal or External) @nmoneypenny wwww.Synxi.com SharePoint
- 29. Content, subject and people recommendations sourced from tibbr Recommended cross-contextualized and personalized SharePoint documents @nmoneypenny wwww.Synxi.com
- 30. Context Aware Personalizing and delivering what’s most relevant to the user’s current activities Personalization Recommendations Machine learningbased inferences of interests and expertise @nmoneypenny wwww.Synxi.com Recommend knowledge and expertise (i.e., content and other users). Engineered serendipity!
Editor's Notes
- #10: There is a new analytics platform in SharePoint 2013 that completely replaces the Web Analytics service application from SharePoint 2010. We had some very specific reasons why we decided to take this approach. First, there was no ability to do item to item recommendations. For example, users who viewed this item also viewed these three other things. Secondly, it didn’t give us a way to promote search results based on an item’s popularity. This means being able to have items that are viewed more frequently percolate up higher in a set of search results. It also didn’t have a way to account for views of list items – so you couldn’t tell what items in a list were being viewed most frequently. Finally, from a hardware perspective it sometimes required a big server to power the Web Analytics service application, and even at that we hit certain thresholds where there was just more data than we could report on.**********************************************Pg. 121
- #11: The new Analytics Processing features in SharePoint 2013 is designed to resolve those issues with these features. You can modify search relevance based on how frequently an item has been viewed – whether from clicking on a search result or just clicking through an item in a site. You can pull up reports directly in a list or library to see how frequently each items has been viewed, both individually and compared to other items in the library. You can look at discussion threads to see which ones are getting the most views, and you can also add this popularity information to page views you create with the Content By Search web part. The model is also extensible so third parties can add new events and track them using the same platform.********************************Pg. 121
- #12: The analytics data processing process is done by the search service application in SharePoint 2013. Usage data like views and clicks from normal site traffic activity are combined with click through and other search metrics and then pushed in the analytics reporting database. A small piece of that data – the recent and all time view count info – is also pushed into the search index. That’s what gives you the capability to use that usage info when you’re looking at search results. An analytics processing job is responsible for examining data for clicks, links and tags, as well as aggregating all of the usage data, to create that data for the analytics reporting database.That concludes what’s new in analytics in SharePoint 2013 – now let’s look at a demo.****************************************************************************Pg. 122
- #23: The kernel trick comes to rescue
- #27: Additionally if no specific context is required, the most relevant content and expertise recommendations can be added to each users MySite or home page to a create a more sticky, truly personalized experience. Recommendations anticipate what users need.
- #28: Additionally if no specific context is required, the most relevant content and expertise recommendations can be added to each users MySite or home page to a create a more sticky, truly personalized experience. Recommendations anticipate what users need.