Show me the Cache: Optimizing Cache-Friendly Recommendations for Sequential Content Access

Abstract

Caching has been successfully applied in wired networks, in the context of Content Distribution Networks (CDNs), and is quickly gaining ground for wireless systems. Storing popular content at the edge of the network (e.g. at small cells) is seen as a "win-win" for both the user (reduced access latency) and the operator (reduced load on the transport network and core servers). Nevertheless, the much smaller size of such edge caches, and the volatility of user preferences suggest that standard caching methods do not suffice in this context. What is more, simple popularity-based models commonly used (e.g. IRM) are becoming outdated, as users often consume multiple contents in sequence (e.g. YouTube, Spotify), and this consumption is driven by recommendation systems. The latter presents a great opportunity to bias the recommender to minimize content access cost (e.g. maximizing cache hit rates). To this end, in this paper we first propose a Markovian model for recommendation-driven user requests. We then formulate the problem of biasing the recommendation algorithm to minimize access cost, while maintaining acceptable recommendation quality. We show that the problem is non-convex, and propose an iterative ADMM-based algorithm that outperforms existing schemes, and shows significant potential for performance improvement on real content datasets.978-1-5386-4725-7/18/$31.00 c 2018 IEEE

The Role of Caching in Future Communication Systems and Networks

“…There is clearly an opportunity to couple caching policies with the importance of each content file, measured in terms of revenue (user payments). First steps towards this direction have been made, e.g., see [54], [129] where content popularity is not the sole caching criterion. Charging the cached content delivery in proportion to the induced bandwidth consumption, inversely proportional to its expected cache hit ratio, or based on the service quality improvement it offers to the user, are only some first intuitive suggestions worthwhile investigating.…”

Leveraging recommendation systems for improving caching emerging short video in content delivery network

“…This indicator reflects the end‐to‐end efficiency of the method. LRU, LFU, collaborative filtering recommendation algorithm, and recent progresses: ABT, 5 Nudge, 6 CARS, 7 Femto, 8 CABaRet, 10 DLC, 32 and CoStore 2 are used as comparison mechanisms.…”

“…In brief, in existing recommendation‐based approaches, for example, References 6‐8, relative content objects that are cached, and are also recommended to users. However, this requires the system to be aware of the content relations (eg, similarity scores, trending videos), that is, information owned by the content provider.…”

“…A recently proposed solution for increasing the efficiency of CDN is jointly caching and recommending content 6‐8 . Recommendation systems (RS) are leveraged to steer recommendations toward cached contents, which can significantly increase the cache hit ratio, even with short cache space.…”

Temporal-Spatial Recommendation for Caching at Base Stations via Deep Reinforcement Learning

“…In [22], the cached files will be recommended if the originally requested file is not in cache, and the caching policy was optimized for BS to improve a ''soft cache hits'' ratio. In [23], some cached files are recommended to a user despite that the cached files may not be preferred by the user, and the recommendation policy in one time slot was optimized. In [24], files are proactively pushed to users, and personalized recommendation is integrated with proactive pushing to shape user demand in order to improve the cache efficiency.…”