# CoGSeq-LTV: Scalable Missing-Aware Graph-Sequence Modeling for Multi-Business Transaction Analytics

DevFeed: [CoGSeq-LTV: Scalable Missing-Aware Graph-Sequence Modeling for Multi-Business Transaction Analytics](<https://devfeed.tech/articles/cogseq-ltv-scalable-missing-aware-graph-sequence-modeling-for-multi-business-transaction-analytics-115218.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2611.01817>)

Author: Ziyang Su, Yao Zhou, Kelin Liu, Guanjie Zheng, Linghe Kong

Published: 2026-11-03T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>)

Topics: [Functional programming](<https://devfeed.tech/topics/functional-programming.md>), [Graph Mining](<https://devfeed.tech/topics/graph-mining.md>), [product analytics](<https://devfeed.tech/topics/product-analytics.md>)

Tags: [alignment](<https://devfeed.tech/tags/alignment.md>), [analytics](<https://devfeed.tech/tags/analytics.md>), [batch](<https://devfeed.tech/tags/batch.md>), [business](<https://devfeed.tech/tags/business.md>), [context](<https://devfeed.tech/tags/context.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [data](<https://devfeed.tech/tags/data.md>), [dataset](<https://devfeed.tech/tags/dataset.md>), [evaluation](<https://devfeed.tech/tags/evaluation.md>), [future](<https://devfeed.tech/tags/future.md>), [global](<https://devfeed.tech/tags/global.md>), [graph](<https://devfeed.tech/tags/graph.md>), [handle](<https://devfeed.tech/tags/handle.md>), [handling](<https://devfeed.tech/tags/handling.md>), [horizon](<https://devfeed.tech/tags/horizon.md>), [large](<https://devfeed.tech/tags/large.md>), [logs](<https://devfeed.tech/tags/logs.md>), [modeling](<https://devfeed.tech/tags/modeling.md>), [sequence](<https://devfeed.tech/tags/sequence.md>), [transaction](<https://devfeed.tech/tags/transaction.md>), [transaction-analytics](<https://devfeed.tech/tags/transaction-analytics.md>)

## AI overview

CoGSeq-LTV combines a batch-induced user-item graph, business-aware transaction sequence encoders, and cross-view contrastive alignment to predict customer value across businesses and time horizons. Reserved unknown-item handling accommodates missing item identifiers, while the sequence encoders retain recency, event gaps, payment context, and goods-table availability. For sparse, zero-inflated, heavy-tailed cumulative targets, it uses a zero-inflated mixture log-normal output with MAE-aligned point supervision and validation-calibrated zero-aware inference. On the BiBiSheng dataset, the authors report improvements over representative LTV baselines across businesses and horizons, including in a later out-of-time evaluation window; the excerpt provides no metric details.

## Source excerpt

arXiv:2611.01817v1 Announce Type: new Abstract: Large-scale super-apps generate heterogeneous transaction logs in which users interact with multiple businesses, item identifiers are only partially available, and future value targets are sparse, heavy-tailed, and cumulative over time. We study multi-business, multi-horizon value prediction as a data-intensive analytics problem over such logs, rather than as a single-table task. (i) a batch-induced user-item graph that provides collaborative signals without materializing a global graph and remains applicable when item identifiers are missing through reserved unknown-item handling, (ii) business-aware sequence encoders that preserve recency, inter-event gaps, payment context, and goods-table availability, and (iii) cross-view contrastive alignment that transfers user-level intent across businesses while retaining business-specific temporal patterns. To handle zero-inflated and long-tailed cumulative targets, CoGSeq-LTV adopts a zero-inflated mixture log-normal head with MAE-aligned point supervision and validation-calibrated zero-aware inference. Experiments on the BiBiSheng transaction-log dataset show that CoGSeq-LTV improves over representative LTV baselines across businesses and horizons, including under a later out-of-time evaluation window.