Beyond Explicit and Implicit: How Users Provide Feedback to Shape Personalized Recommendation Content

TL;DR

This study investigates how users provide feedback to personalize algorithmic recommendation feeds on social platforms, revealing explicit, intentional implicit, and unintentional implicit feedback as key signals. Through 34 semi-structured interviews on Xiaohongshu and Douyin, it demonstrates that feedback type choices align with four user purposes: content consumption, directed information seeking, content creation/promotion, and feed customization. A central contribution is the introduction of intentional implicit feedback, capturing deliberate actions users take to influence learning signals without direct input, which has implications for user agency and algorithm transparency. The findings offer design guidance to support intention-aware feedback, purpose-oriented controls, and privacy-conscious data collection in recommender systems, with practical impact for platform design and user experience.

Abstract

As personalized recommendation algorithms become integral to social media platforms, users are increasingly aware of their ability to influence recommendation content. However, limited research has explored how users provide feedback through their behaviors and platform mechanisms to shape the recommendation content. We conducted semi-structured interviews with 34 active users of algorithmic-driven social media platforms (e.g., Xiaohongshu, Douyin). In addition to explicit and implicit feedback, this study introduced intentional implicit feedback, highlighting the actions users intentionally took to refine recommendation content through perceived feedback mechanisms. Additionally, choices of feedback behaviors were found to align with specific purposes. Explicit feedback was primarily used for feed customization, while unintentional implicit feedback was more linked to content consumption. Intentional implicit feedback was employed for multiple purposes, particularly in increasing content diversity and improving recommendation relevance. This work underscores the user intention dimension in the explicit-implicit feedback dichotomy and offers insights for designing personalized recommendation feedback that better responds to users' needs.

Figures (2)

• Figure 1: The main user interfaces of Xiaohongshu and Douyin. (\ref{['fig:feedback']}) is the main page of Xiaohongshu, the "Explore" page, displaying a selection of posts recommended by the algorithm, consisting of both picture and video posts. Located at the top of the main page is the search bar. Below the search bar, users often see trending hashtags and their interested channels that can be customized. Long press on a post can trigger options for reporting a post, including "Not interested" and "Content feedback." (\ref{['fig:xhspost']}) is a note detail page, where the note itself is the centerpiece. It typically includes a mix of text, images (or videos), and hashtags. Users can follow the creator, like, collect, or leave comments on the post. (\ref{['fig:douyinmain']}) is the main page of Douyin, the "For You" page, showcasing a continuous stream of short videos curated by Douyin's algorithm for each user. A new video automatically displays as you scroll (or swipe) vertically. Users can like a video, leave a comment, share it, or follow the creator using the buttons on the right side. Douyin also provides similar content feedback features.
• Figure 2: The main user interfaces of Kuaishou and Bilibili Shorts. (\ref{['fig:ksmain']}) is the main page of Kuaishou, the "Trend" page, which displays a selection of posts recommended by the algorithm. The feed consists mostly of video posts but also includes picture posts, indicated by a picture icon in the top-left corner. This page is similar to the Xiaohongshu "Explore" page in Figure \ref{['fig:feedback']}. (\ref{['fig:ksvideo']}) is a short video post page of Kuaishou. Once entering this page, users can scroll vertically to switch video posts, similar to Douyin's "For You" page. (\ref{['fig:bili']}) is the main interface of Bilibili Shorts, similar to the main interface of Douyin.