---
title: "Does Every Data Instance Matter? Enhancing Sequential Recommendation by Eliminating Unreliable Data"
authors:
  - "Yatong Sun"
  - "Bin Wang"
  - "Zhu Sun"
  - "Xiaochun Yang"
doi: "10.24963/ijcai.2021/218"
url: "https://ijcai.org/proceedings/2021/218"
pdf: "https://ijcai.org/proceedings/2021/0218.pdf"
venue: "Proceedings of the Thirtieth International Joint Conference on Artificial Intelligence"
year: 2021
pages: "1579-1585"
track: "Main Track"
keywords:
  - "Humans and AI: Personalization and User Modeling"
  - "Data Mining: Recommender Systems"
---

# Does Every Data Instance Matter? Enhancing Sequential Recommendation by Eliminating Unreliable Data

Yatong Sun, Bin Wang, Zhu Sun, Xiaochun Yang

*Proceedings of the Thirtieth International Joint Conference on Artificial Intelligence*, pages 1579-1585. DOI: [10.24963/ijcai.2021/218](https://doi.org/10.24963/ijcai.2021/218)

## Abstract

Most sequential recommender systems (SRSs) predict next-item as target for each user given its preceding  items  as  input,  assuming  that  each  input is  related  to  its  target.   However,  users  may  unintentionally  click  on  items  that  are  inconsistent with  their  preference.   We  empirically  verify  that SRSs  can  be  misguided  with  such  unreliable  instances  (i.e.   targets  mismatch  inputs).   This  inspires  us  to  design  a  novel  SRS By Eliminating unReliable Data (BERD) guided with two observations:  (1) unreliable instances generally have high training  loss;  and  (2)  high-loss  instances  are  not necessarily  unreliable  but  uncertain  ones  caused by blurry sequential pattern.  Accordingly, BERD models both loss and uncertainty of each instance via a Gaussian distribution to better distinguish unreliable instances; meanwhile an uncertainty-aware graph  convolution  network  is  exploited  to  assist in mining unreliable instances by lowering uncertainty.   Extensive  experiments  on  four  real-world datasets  demonstrate  the  superiority  of  our  proposed BERD.