The Retrieval-Augmented Language Model (RALM) has demonstrated remarkable performance on knowledge-intensive tasks by integrating external knowledge during inference, which mitigates the factual hallucinations inherited in large language models (LLMs). Despite these advancements, challenges persist in the implementation of RALMs, particularly in terms of reliability and traceability. Specifically, the irrelevant document retrieval may result in unhelpful responses or even deteriorate the performance of LLMs, while the lack of appropriate citations in outputs complicates efforts to verify the trustworthiness of the models. To this end, we propose a novel self-reasoning framework aimed at improving the reliability and traceability of RALMs, whose core idea is to leverage reasoning trajectories generated by the LLM itself. The framework involves constructing self-reasoning trajectories through three processes: a relevance-aware process, an evidence-aware selective process, and a trajectory analysis process. We evaluated our framework across four public datasets (two short-form QA datasets, one long-form QA dataset, and one fact verification dataset) to demonstrate its superiority. Our method can outperform existing state-of-the-art models and achieve performance comparable with GPT-4, using only 2,000 training samples.

Yuan Xia, Zhenhui Shi, Jingbo Zhou, Jiayu Xu, Chao Lu, Yehui Yang, Lei Wang, Haifeng Huang, Xia Zhang, Junwei Liu. Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing. 2022.

Medical entity relation verification is a crucial step to build a practical and enterprise medical knowledge graph (MKG) because high-precision medical entity relation is a key requirement for many MKG-based applications. Existing relation verification approaches for general knowledge graphs are not designed for considering medical domain knowledge, although it is central to achieve high-quality entity relation verification for MKG. To this end, in this paper, we introduce a system for medical entity relation verification with large-scale machine reading comprehension. The proposed system is tailored to overcome the unique challenges of medical relation verification including high variants of medical terms, the high difficulty of evidence searching in complex medical documents, and the lack of evidence labels for supervision. To deal with the problem of variants of medical terms, we introduce a synonym-aware retrieve model to retrieve the potential evidence implicitly verifying the given claim. To better utilize the medical domain knowledge, a relation-aware evidence detector and a medical ontology-enhanced aggregator are developed to improve the performance of the relation verification module. Moreover, to overcome the challenge of providing high-quality evidence due to the lack of labels, we introduce an interactive collaborative-training method to iteratively improve the evidence accuracy. Finally, we conduct extensive experiments to demonstrate that the performance of our proposed system is superior to all comparable models. We also demonstrate that our system can significantly reduce the annotation time by medical experts in real-world verification tasks. It can help to improve the efficiency by nearly 300%. In particular, our system has been embedded into the Baidu Clinical Decision Support System.

Automatic diagnosis systems have attracted increasing attention in recent years. The reinforcement learning (RL) is an attractive technique for building an automatic diagnosis system due to its advantages for handling sequential decision making problem. However, the RL method still cannot achieve good enough prediction accuracy. In this paper, we propose a Generative Adversarial regularized Mutual information Policy gradient framework (GAMP) for automatic diagnosis which aims to make a diagnosis rapidly and accurately. We first propose a new policy gradient framework based on the Generative Adversarial Network (GAN) to optimize the RL model for automatic diagnosis. In our framework, we take the generator of GAN as a policy network, and also use the discriminator of GAN as a part of the reward function. This generative adversarial regularized policy gradient framework can try to avoid generating randomized trials of symptom inquires deviated from the common diagnosis paradigm. In addition, we add mutual information to enhance the reward function to encourage the model to select the most discriminative symptoms to make a diagnosis. Experiment evaluations on two public datasets show that our method beats the state-of-art methods, not only can achieve higher diagnosis accuracy, but also can use a smaller number of inquires to make diagnosis decision.

Medical Question Answering (medical QA), which studies the problem of automatically answering patients' medical questions online, is one of the major applications of bioinformatics. Though many efforts have been made before, the medical QA system still deserves delicate algorithm optimization due to the serious application scenario and strict requirement for the answer quality. In this paper, we introduce a novel Knowledge Abstraction Matching (KAM) method for the medical QA problem. The intuition of KAM is that there are many frequent repeat text segments appearing in the answers across different questions. From this view, we propose a new method that consists of frequent segment$N$-gram mining, medical knowledge abstraction, medical segment matching and answer re-retrieval. KAM has been incorporated into Baidu's enterprise medical QA system MelodyQA deployed on the backend of Muzhi Doctor. The evaluation shows that the proposed method can generate more quality answers for MelodyQA with a significant improvement of question coverage under acceptable accuracy.

Vessel enhancement is an important pre-processing step of applications in vessel image analysis. However, most of the current methods are developed merely based on the intensity variety inside and outside vessel instead of considering the vessel path, which emphasizes the vascular structures via characterizing additional connectivity and length information. Aiming at further utilizing beneficial length information of vessels, we propose a novel method to impose length constraint on Hessian information for vessel enhancement. Specifically, Eigen analysis of multiscale Hessian matrix has been taken at each pixel for the local vesselness response and direction information. Then, vessel path is searched along each pixel's direction, as well as maintains the property of curvilinear smoothness. The proposed method is compared with three conventional vessel enhancement methods. The experiment results show that our proposed approach has the advantages of the fine response of low-contrast vessel region and less noise background. In addition, the quantity evaluation indicates that a state-of-art vessel enhancement performance could be achieved compared with other methods.