Despite the recognized potential of multimodal data to improve model accuracy, many large-scale industrial recommendation systems, including Taobao display advertising system, predominantly depend on sparse ID features in their models. In this work, we explore approaches to leverage multimodal data to enhance the recommendation accuracy. We start from identifying the key challenges in adopting multimodal data in a manner that is both effective and cost-efficient for industrial systems. To address these challenges, we introduce a two-phase framework, including: 1) the pre-training of multimodal representations to capture semantic similarity, and 2) the integration of these representations with existing ID-based models. Furthermore, we detail the architecture of our production system, which is designed to facilitate the deployment of multimodal representations. Since the integration of multimodal representations in mid-2023, we have observed significant performance improvements in Taobao display advertising system. We believe that the insights we have gathered will serve as a valuable resource for practitioners seeking to leverage multimodal data in their systems.

Click-through rate (CTR) prediction plays as a core function module in various personalized online services. The traditional ID-based models for CTR prediction take as inputs the one-hot encoded ID features of tabular modality, which capture the collaborative signals via feature interaction modeling. But the one-hot encoding discards the semantic information conceived in the original feature texts. Recently, the emergence of Pretrained Language Models (PLMs) has given rise to another paradigm, which takes as inputs the sentences of textual modality obtained by hard prompt templates and adopts PLMs to extract the semantic knowledge. However, PLMs generally tokenize the input text data into subword tokens and ignore field-wise collaborative signals. Therefore, these two lines of research focus on different characteristics of the same input data (i.e., textual and tabular modalities), forming a distinct complementary relationship with each other. In this paper, we propose to conduct Fine-grained feature-level ALignment between ID-based Models and Pretrained Language Models (FLIP) for CTR prediction. We design a novel joint reconstruction pretraining task for both masked language and tabular modeling. Specifically, the masked data of one modality (i.e., tokens or features) has to be recovered with the help of the other modality, which establishes the feature-level interaction and alignment via sufficient mutual information extraction between dual modalities. Moreover, we propose to jointly finetune the ID-based model and PLM for downstream CTR prediction tasks, thus achieving superior performance by combining the advantages of both models. Extensive experiments on three real-world datasets demonstrate that FLIP outperforms SOTA baselines, and is highly compatible for various ID-based models and PLMs.