Browse Prior Art Database

Mapping advisors to users

IP.com Disclosure Number: IPCOM000253947D
Publication Date: 2018-May-16
Document File: 5 page(s) / 64K

Publishing Venue

The IP.com Prior Art Database

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 38% of the total text.

Mapping advisors to users

Most software product users begin as novices and require a lot of learning before becoming experts. During that learning period, users could benefit from having a tutor accessible, especially when the complexity of tasks increases. It would also be helpful for said tutor to have information about the student’s history and personality, to effectively communicate and optimize support. In addition, expertise is spread across many different people and organizations, which reduces accessibility. Users need support from domain experts that are external to the company that developed the software. Many applications have a product specific help/support system in place, but these mostly describe the functionality in general terms (e.g., do this step first, then click on that menu, change that setting, etc.). This type of expert advice does not consider individual traits, background, experience, and to whom it is best for the user to speak for help in a particular context. The novel solution is a mapping advisors system. This system matches a software user with an expert advisor for the current task. This system ensures that the user receives personalized support based on individual personality and knowledge needs. To determine the appropriate matches, the system considers:

 The user’s profile: personality, usage history, task context, transactional data, and current applications in use

 The expert advisor’s profile and content: personality, domain knowledge, and content

The result is a list of the most appropriate and available experts from a federated group of advisors and applicable content from those advisors that match the user’s profile. When interactions between users and advisors begin (i.e., selecting content, chatting with the advisor, etc.) the system continues to update the profile of the user for future improvement of the matching algorithm. Expert Advisor determines the context via analysis of the displayed text from the calling application as computed via a custom and Expert Advisor-specific machine learning (ML) algorithm that incorporates a feedback loop to improve context recognition based on instrumentation that measures if users refine Expert Advisor suggestions after the initial set of suggestions by product, topic within the product, computed user expertise level and other data points. The computed context is used via a second Expert Advisor-specific ML algorithm that

provides the initial set of advisors based on associated profile information and public (can include private based on deployment) articles/books/blogs/videos, while continuously updating and improving the list based on elements such as advisor feedback ratings, participation, products, and topics of engagement. Expert Advisor is product agnostic with a pluggable architecture for chat and search, thus allowing its use with products that are already used by an organizational entity while also allowing its use on a public or private network. T...