Why listed? One of the next steps with my own Vendelligence information support system project is building up the local query repository and then querying it with native and external search functions. I plan to plug into existing Machine Learning APIs as a starting point once I have a reasonably usable data set. Before I go using any API, I need to understand the concepts and how it works to use it effectively. This article is part of a series, which ties in with my self-learning objectives in the coming months. I will need to at least be able to understand terms if I am asking a colleague or a friend with expertise in such domains for help, so we can have a meaningful and productive discussion.
Source 1: Google Smart Reply Research Paper
Why listed? I am adding two links from Adrian Colyer this week from The Morning Paper. I found his blog earlier this year and it was like somebody had finally delivered a universal translator for some of those barely readable computer science papers at the end of ACM magazine or in online publications – he intentionally tries to avoid overly mathematical computer science papers to keep the blog digestible. In general, however, Adrian takes complex computer science research data and puts it in a language you can understand. Adrian is providing a wonderful knowledge service here in making these papers accessible to a wider audience. The first one picked this week covers the design of the Google Smart Reply mail service. The other discusses the Microsoft Research on achieving human parity in conversational speech and speech recognition is a personal interest area with information systems in group meeting contexts rather than solo machine interaction scenarios for myself.
Why listed? I am moving onto workflow modelling, local query testing, storage, and retrieval of information via a local private build of Vendelligence over the past week. After adding more vendors to the private edition in recent days, I started hitting both workflow automation and user interface challenges, as well as identifying future system limits based on the number of links per vendor or open source project. I needed to look to teams who have faced information system and human augmentation challenges to see how they solved them and instead of looking at current buzzword articles, I decided to look back a few decades into the past for help. Certain names kept cropping up in online videos and personal reading, so I decided to write about two individuals that stood out to me and why I found their work inspiring in general.