CogSearch: A Cognitive-Aligned Multi-Agent Framework for Proactive Decision Support in E-Commerce Search DOI: XXXXXXX.XXXXXXX Conference: Make sure to enter the correct conference title from your rights confirmation email; June 03–05, 2018; Woodstock, NY ISBN: 978-1-4503-XXXX-X/2018/06 CCS: Computing methodologies Multi-agent systems CCS: Information systems Novelty in information retrieval

Zhouwei Zhai, Mengxiang Chen, Haoyun Xia, Jin Li, Renquan Zhou, Min Yang · 2026

Read CogSearch: A Cognitive-Aligned Multi-Agent Framework for Proactive Decision Support in E-Commerce Search DOI: XXXXXXX.XXXXXXX Conference: Make sure to enter the correct conference title from your rights confirmation email; June 03–05, 2018; Woodstock, NY ISBN: 978-1-4503-XXXX-X/2018/06 CCS: Computing methodologies Multi-agent systems CCS: Information systems Novelty in information retrieval by Zhouwei Zhai, Mengxiang Chen, Haoyun Xia, Jin Li, Renquan Zhou, Min Yang (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2603-11927.txt.

Introduction

Zhai and co-authors present CogSearch, a multi-agent framework that turns e-commerce search into proactive decision support. Standard search just retrieves and ranks, leaving shoppers to juggle hard choices alone. Four specialist agents split the work: breaking down user intent, fusing store and outside knowledge, and shaping clear answers. Offline tests favor it on consultative and complex queries. Live A/B tests on JD.com cut decision costs by 5% and lifted overall conversion slightly. For decision-heavy queries, conversion jumped by 30%.

Catalog record

CREATORZhouwei Zhai, Mengxiang Chen, Haoyun Xia, Jin Li, Renquan Zhou, Min Yang
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE21548
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2603.11927

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

CogSearch: A Cognitive-Aligned Multi-Agent Framework for Proactive Decision Support in E-Commerce Search DOI: XXXXXXX.XXXXXXX Conference: Make sure to enter the correct conference title from your rights confirmation email; June 03–05, 2018; Woodstock, NY ISBN: 978-1-4503-XXXX-X/2018/06 CCS: Computing methodologies Multi-agent systems CCS: Information systems Novelty in information retrieval Zhouwei Zhai, Mengxiang Chen, Haoyun Xia, Jin Li, Renquan Zhou, Min Yang Abstract . Modern e-commerce search engines, largely rooted in passive retrieval-and-ranking models, frequently fail to support complex decision-making, leaving users overwhelmed by cognitive friction. In this paper, we introduce CogSearch, a novel cognitive-oriented multi-agent framework that reimagines e-commerce search as a proactive decision support system. By synergizing four specialized agents, CogSearch mimics human cognitive workflows: it decomposes intricate user intents, fuses heterogeneous knowledge across internal and external sources, and delivers highly actionable insights. Our offline benchmarks validate CogSearch’s excellence in consultative and complex search scenarios. Extensive online A/B testing on JD.com demonstrates the system’s transformative impact: it reduced decision costs by 5% and achieved a 0.41% increase in overall UCVR, with a remarkable 30% surge in conversion for decision-heavy queries. CogSearch represents a fundamental shift in information retrieval, moving beyond traditional relevance-centric paradigms toward a future of holistic, collaborative decision intelligence. 1. Introduction E-commerce search serves as the core hub connecting user needs with product supply. Its fundamental objective extends beyond merely returning relevant products; it aims to assist users throughout the entire cognitive process from “need articulation” to “decision formation.” Industrial search systems have long evolved around the “retrieval-ranking” paradigm. The retrieval stage has progressed from statistical models like LSI ( 4 ) , BM25 ( 12 ) , and TF-IDF ( 1 ) to deep semantic matching driven by approaches like RocketQA ( 11 ) . The ranking stage enhances precision through techniques such as knowledge distillation ( 9 ) , DIN attention mechanisms ( 20 ) , and personalized modeling ( 18 ) , forming a closed-loop optimization process ( 14 ) . Recently, Large Language Models have further augmented system performance via query rewriting ( 3 ; 10 ) , generative retrieval ( 16 ; 2 ; 8 ) , and relevance fine-tuning ( 17 ; 15 ; 5 ) . However, existing research still faces critical bottlenecks: the passive response mode hinders the system’s ability to handle complex, multi-attribute implicit reasoning needs, and reliance solely on platform-internal data limits cross-platform information integration. More crucially, systems often function as “result displays” rather than “decision facilitators,” leading to excessive cognitive load and comparison costs for users during high-complexity decision-making. The core challenge in addressing these issues lies in breaking the limitations of static indexing to achieve deep convergence of user intent within dynamic dialogues and transforming fragmented product attributes into structured decision-making bases. Inspired by Simon’s decision-making model ( 13 ) , this paper proposes CogSearch, a cognitively-aligned multi-agent framework. This framework upgrades search from “keyword matching” to “active decision assistance” through the collaborative operation of four specialized agents: Planner, Executor, Guider, and Decider. Our core contributions are: • Architectural Innovation: We propose the CogSearch multi-agent framework. The Planner decomposes complex intents, the Executor integrates multi-source information, the Guider steers demand convergence, and the Decider provides structured purchasing recommendations. This achieves precise alignment between the search process and the user’s shopping cognitive chain. • Large-Scale Industrial Validation: CogSearch has been fully deployed on JD.com, one of China’s largest self-operated e-commerce platforms. Offline experiments demonstrate significant performance gains in handling complex queries. Furthermore, online A/B testing confirms that our framework reduces user decision costs by 5%, yields a 0.41% lift in overall User Conversion Rate (UCVR), and achieves a remarkable 30% UCVR increase for decision-heavy queries. 2. The CogSearch Framework 2.1. System Overview Inspired by Simon’s three-stage decision-making model ( 13 ) , CogSearch fundamentally redefines e-commerce search as a collaborative cognitive decision-making process, as opposed to the traditional retrieval-ranking pipeline, as illustrated in Figure 1 . Figure 1. CogSearch Framework Overview The design logic of CogSearch deeply aligns with the user’s shopping cognitive chain: • Planner Agent: Solves the problem of demand understanding. Based on the user’s current query, search history, and historical click context, it comprehends user intent and performs task planning, generating a task graph stored in the Memory System. Specific tasks include product search tasks, web search tasks, and tool/service invocation. • Executor Agent: Solves the problem of multi-source information gathering. Based on the Planning Agent’s output, it executes multi-source retrieval tasks – encompassing in-platform product retrieval (returning a list of eligible products), web-scale information retrieval, and tool/service invocation (e.g., weather, logistics APIs). • Guider Agent: Solves the problem of proactive decision guidance. Leveraging retrieved product results combined with web-sourced information, it generates personalized dynamic filters and purchasing strategies. Furthermore, it proactively identifies latent needs to generate follow-up questions for demand convergence or stimulate new user requirements. • Decider Agent: Solves the problem of integrated decision-making. Synthesizing the user profile, historical behavioral context, interaction signals from the Navigation Agent (e.g., filter clicks), and multi-source information (product attributes + web guides/strategies), it employs LLM-based reasoning to output final product recommendations accompanied by structured decision rationale. Memory System acts as the hub for storing, updating, and maintaining all contextual information for the agents. This includes contextual memories, agent states, and the task graph. 2.2. Planner Agent: Intent-Driven Task Generation As the cognitive orchestration engine of CogSearch, the Planner Agent transcends traditional query rewriting by formulating the search process as a structured reasoning problem. Its primary objective is to bridge the semantic gap between ambiguous user requirements and precise system execution capabilities through the generation of an executable Task Graph. Formally, we define the planner’s input as a dynamic context tuple: (1) \mathcal{C}_{t}=\{q_{t},\mathcal{H}^{search}_{<t},\mathcal{H}^{click}_{<t}\} where q_{t} denotes the current user query, while \mathcal{H}^{search}_{<t} and \mathcal{H}^{click}_{<t} represent the historical query sequence and item interaction context, respectively. The Planner leverages the reasoning capacity of Large Language Models to map \mathcal{C}_{t} into a Directed Acyclic Graph (DAG), denoted as: \mathcal{G}=(\mathcal{V},\mathcal{E}) . The vertex set \mathcal{V}=\{v_{1},v_{2},...,v_{n}\} represents atomic cognitive tasks derived from the user’s latent intent. To facilitate comprehensive decision support, we define the action space of \mathcal{V} across three functional dimensions: (1) ProductSearch ( T_{prod} ) for retrieving candidate items from the internal E-commerce inventory; (2) WebSearch ( T_{web} ) for acquiring open-world knowledge (e.g., reviews, usage guides) to support cross-platform information fusion; and (3) ToolInvocation ( T_{tool} ) for executing specific functional operations via the Model Context Protocol (MCP), such as price comparison or logistics querying. The edge set \mathcal{E} explicitly models the semantic dependencies between these tasks. A directed edge (v_{i},v_{j})\in\mathcal{E} is generated if and only if the execution of task v_{j} is conditional on the output context of v_{i} , satisfying the constraint Input(v_{j})\subseteq Output(v_{i}) . The generation of \mathcal{G} is modeled as a conditional probability distribution P_{\theta}(\mathcal{G}|\mathcal{C}_{t}) , where the Planner decomposes complex queries into a topological sequence of sub-tasks. 2.3. Executor Agent: Multi-Source Acquisition 2.3.1. ProductSearch Agent This agent transforms structured queries into ranked product candidates via a cognitively augmented hybrid retrieval architecture. To balance precision and diversity: 1) Keyword-Vector Fusion: Concurrently invokes platform APIs (lexical recall) and contrastive learning-based semantic embedding models (intent generalization). 2) Multi-Source Enrichment: Coordinates ProductAttr and ReviewSum sub-agents to aggregate attributes and

Read the complete text