Addressing Item Cold Start Problem in Collaborative Filtering-Based Recommender Systems Using Auxiliary Information

Abstract

The initial interaction of a user with a recommender system is problematic because, in such a so-called cold start situation, the recommender system has very little information about the user, if any. Moreover, in collaborative filtering, users need to share their preferences with the service provider by rating items while in content-based filtering there is no need for such information sharing. We have recently shown that a content-based model that uses hypercube graphs can determine user preferences with a very limited number of ratings while better preserving user privacy. In this paper, we confirm these findings on the basis of experiments with more than 1,000 users in the restaurant and movie domains. We show that the proposed method outperforms standard machine learning algorithms when the number of available ratings is at most 10, which often happens, and is competitive with larger training sets. In addition, training is simple and does not require large computational efforts.

Long-Term Preference Mining With Temporal and Spatial Fusion for Point-of-Interest Recommendation

“…V c For the purpose, we transform the user-id, v id and the encoded category information into continuous vectors u u t , v u t and vc u i respectively through embedding layer. Then, to learn the non-linear dependencies in user-POI and category data, we feed the embedding layers to LSTM as in Eq (8). The hidden vector h u t is the representation for user u at step t and captures the POI category preferences of the user over the long term.…”

“…The dearth of explicit ratings and limited physical accessibility aggravates the issue of data sparsity. The sparse check-in matrix aggravates the cold-start problem [8]. POI recommendation must cater to the needs of the cold-start users, and the recommended POIs must involve the POIs that are new or have limited rating data, termed as cold-start POIs.…”

Bridging the Information Gap with Buddy Network: Strategies for Cold Start in Recommender Systems

“…Explicit feedback, on the other hand, involves users assigning discrete ratings to each purchased product. Users and products can be classified as either "warm" (where previous interactions are known) or "cold" [9] (where previous interactions are unknown, often involving new users or products without an interaction history). Recommendations for the latter group present the well-recognized challenge of cold-start recommendations.…”