# EviRec: Continual Evidence Learning for Dual Cold-Start POI Recommendation

arXiv: https://arxiv.org/abs/2609.20313 · HTML: https://mlsanity.com/paper/2609.20313/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 18 of its 32 references that are in the map.
- First posted on arXiv 31 Jul 2026.

## Builds on (most cited first, 18 of 18)

- [Outrageously Large Neural Networks: The Sparsely-Gated Mixture-of-Experts Layer](https://mlsanity.com/paper/1701.06538/) (2017)
- [Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity](https://mlsanity.com/paper/2101.03961/) (2021)
- [Gradient Episodic Memory for Continual Learning](https://mlsanity.com/paper/1706.08840/) (2017)
- [Self-Attentive Sequential Recommendation](https://mlsanity.com/paper/1808.09781/) (2018)
- [LightGCN: Simplifying and Powering Graph Convolution Network for Recommendation](https://mlsanity.com/paper/2002.02126/) (2020)
- [BERT4Rec: Sequential Recommendation with Bidirectional Encoder Representations from Transformer](https://mlsanity.com/paper/1904.06690/) (2019)
- [MeLU: Meta-Learned User Preference Estimator for Cold-Start Recommendation](https://mlsanity.com/paper/1908.00413/) (2019)
- [STAN: Spatio-Temporal Attention Network for Next Location Recommendation](https://mlsanity.com/paper/2102.04095/) (2021)
- [Improved Recurrent Neural Networks for Session-based Recommendations](https://mlsanity.com/paper/1606.08117/) (2016)
- [Warm Up Cold-start Advertisements: Improving CTR Predictions via Learning to Learn ID Embeddings](https://mlsanity.com/paper/1904.11547/) (2019)
- [Large Language Models for Next Point-of-Interest Recommendation](https://mlsanity.com/paper/2404.17591/) (2024)
- [Dynamically Expandable Graph Convolution for Streaming Recommendation](https://mlsanity.com/paper/2303.11700/) (2023)
- [A Diffusion model for POI recommendation](https://mlsanity.com/paper/2304.07041/) (2023)
- [Generative Next POI Recommendation with Semantic ID](https://mlsanity.com/paper/2506.01375/) (2025)
- [GeoGen: A Two-stage Coarse-to-Fine Framework for Fine-grained Synthetic Location-based Social Network Trajectory Generation](https://mlsanity.com/paper/2510.07735/) (2025)
- [GPT4Rec: Graph Prompt Tuning for Streaming Recommendation](https://mlsanity.com/paper/2406.08229/) (2024)
- [REPLAY: Modeling Time-Varying Temporal Regularities of Human Mobility for Location Prediction over Sparse Trajectories](https://mlsanity.com/paper/2402.16310/) (2024)
- [Context-Adaptive Graph Neural Networks for Next POI Recommendation](https://mlsanity.com/paper/2506.10329/) (2025)

## Abstract (arXiv)

Point-of-Interest (POI) recommendation is a core task in location-based services, yet most existing methods assume a fixed user population and POI catalog. Through a large-scale data-driven analysis of 10 U.S. cities, we identify substantial POI churn, user turnover, category drift, and decay in static POI memory, motivating the study of continual dual cold-start POI recommendation. To address this setting, we propose EviRec, a continual evidence-learning framework that estimates how much historical evidence should be trusted separately for each candidate POI. EviRec scores each visible candidate from three complementary views: a matching view based on the user's recent mobility profile, a transition-memory view that captures repeated mobility routines, and a lifecycle view that reflects candidate maturity. Because a near-zero transition score may indicate either irrelevance or insufficient observation, EviRec qualifies the evidence using each candidate's observation state and applies a reliability gate to adaptively route between transition-memory and lifecycle evidence. We evaluate EviRec on a full-year, five-city POI check-in dataset containing more than 30,000 users and 684,200 trajectories. Experimental results show that EviRec consistently outperforms state-of-the-art baselines, with the largest gains concentrated on cold-start queries. In particular, EviRec improves NDCG@10 by 20.4% on Dual-New cases over the strongest baseline. In-depth analyses further confirm that these gains arise primarily from candidate-specific reliability gating while largely preserving previously learned mobility routines.