Recommendation system serves as a conduit connecting users to an incredibly large, diverse and ever growing collection of contents. In practice, missing information on fresh (and tail) contents needs to be filled in order for them to be exposed and discovered by their audience. We here share our success stories in building a dedicated fresh content recommendation stack on a large commercial platform. To nominate fresh contents, we built a multi-funnel nomination system that combines (i) a two-tower model with strong generalization power for coverage, and (ii) a sequence model with near real-time update on user feedback for relevance. The multi-funnel setup effectively balances between coverage and relevance. An in-depth study uncovers the relationship between user activity level and their proximity toward fresh contents, which further motivates a contextual multi-funnel setup. Nominated fresh candidates are then scored and ranked by systems considering prediction uncertainty to further bootstrap content with less exposure. We evaluate the benefits of the dedicated fresh content recommendation stack, and the multi-funnel nomination system in particular, through user corpus co-diverted live experiments. We conduct multiple rounds of live experiments on a commercial platform serving billion of users demonstrating efficacy of our proposed methods.

Article Free Access Share on Rednet: a wireless ATM local area network using infrared links Authors: J. H. Condon AT&T Bell Laboratories, Murray Hill NJ AT&T Bell Laboratories, Murray Hill NJView Profile , T. S. Duff AT&T Bell Laboratories, Murray Hill NJ AT&T Bell Laboratories, Murray Hill NJView Profile , M. F. Jukl AT&T BelI Laboratories, Murray Hill NJ AT&T BelI Laboratories, Murray Hill NJView Profile , C. R. Kalmanek AT&T Bell Laboratories, Murray Hill NJ AT&T Bell Laboratories, Murray Hill NJView Profile , B. N. Locanthi AT&T Bell Laboratories, Murray Hill NJ AT&T Bell Laboratories, Murray Hill NJView Profile , J. P. Savicki AT&T Bell Laboratories, Murray Hill NJ AT&T Bell Laboratories, Murray Hill NJView Profile , J. H. Venutolo AT&T Bell Laboratories, Murray Hill NJ AT&T Bell Laboratories, Murray Hill NJView Profile Authors Info & Claims MobiCom '95: Proceedings of the 1st annual international conference on Mobile computing and networkingDecember 1995 Pages 151–159https://doi.org/10.1145/215530.215568Published:01 December 1995Publication History 14citation523DownloadsMetricsTotal Citations14Total Downloads523Last 12 Months31Last 6 weeks3 Get Citation AlertsNew Citation Alert added!This alert has been successfully added and will be sent to:You will be notified whenever a record that you have chosen has been cited.To manage your alert preferences, click on the button below.Manage my AlertsNew Citation Alert!Please log in to your account Save to BinderSave to BinderCreate a New BinderNameCancelCreateExport CitationPublisher SiteeReaderPDF

We wish to determine the effectiveness of some simple hardware for debugging and profiling compiled programs on a conventional processor. The hardware cost is small -- a counter decremented on each instruction that raises an exception when its value becomes zero. With the counter a debugger can provide data watchpoints and reverse execution: a profiler can measure the total instruction cost of a code segment and sample the program counter accurately. Such a counter has been included on a single-board MC68020 workstation, for which system software is currently being written. We will report our progress at the symposium.

Abstract The Blit is an experimental bitmap graphics terminal built for research into interactive computer graphics on the UNIX time‐sharing system. The hardware is inexpensive and the graphics functions are implemented entirely in software. Nevertheless, the graphics performance of the Blit is comparable or superior to some displays with special‐purpose graphics hardware. This paper explains the paradox by referring to some principles of design: the hardware and software should be designed together to complement each other; carefully designed software can outperform infelicitous hardware; and simplicity of design leads to efficiency of execution. These principles are illustrated by examples from the Blit hardware and software and comparisons with other systems.