arXiv subjects

Feng Ji

Publications and source records attributed to Feng Ji.

At least 91 records · Page 5Linked to original sources

arXiv2020

Signal processing with a distribution of graph operators

In this paper, we develop a signal processing framework of a network without explicit knowledge of the network topology. Instead, we make use of knowledge on the distribution of operators on the network. This makes the framework flexible and useful when accurate knowledge of graph topology is unavailable. Moreover, the usual graph signal processing is a special case of our framework by using the delta distribution. The main elements of the theory include Fourier transform, theory of filtering and sampling.

eess.SP↗

arXiv2020

Improving Commonsense Question Answering by Graph-based Iterative Retrieval over Multiple Knowledge Sources

In order to facilitate natural language understanding, the key is to engage commonsense or background knowledge. However, how to engage commonsense effectively in question answering systems is still under exploration in both research academia and industry. In this paper, we propose a novel question-answering method by integrating multiple knowledge sources, i.e. ConceptNet, Wikipedia, and the Cambridge Dictionary, to boost the performance. More concretely, we first introduce a novel graph-based iterative knowledge retrieval module, which iteratively retrieves concepts and entities related to the given question and its choices from multiple knowledge sources. Afterward, we use a pre-trained language model to encode the question, retrieved knowledge and choices, and propose an answer choice-aware attention mechanism to fuse all hidden representations of the previous modules. Finally, the linear classifier for specific tasks is used to predict the answer. Experimental results on the CommonsenseQA dataset show that our method significantly outperforms other competitive methods and achieves the new state-of-the-art. In addition, further ablation studies demonstrate the effectiveness of our graph-based iterative knowledge retrieval module and the answer choice-aware attention module in retrieving and synthesizing background knowledge from multiple knowledge sources.

cs.CL↗

arXiv2020

AliMe KG: Domain Knowledge Graph Construction and Application in E-commerce

Pre-sales customer service is of importance to E-commerce platforms as it contributes to optimizing customers' buying process. To better serve users, we propose AliMe KG, a domain knowledge graph in E-commerce that captures user problems, points of interests (POI), item information and relations thereof. It helps to understand user needs, answer pre-sales questions and generate explanation texts. We applied AliMe KG to several online business scenarios such as shopping guide, question answering over properties and recommendation reason generation, and gained positive results. In the paper, we systematically introduce how we construct domain knowledge graph from free text, and demonstrate its business value with several applications. Our experience shows that mining structured knowledge from free text in vertical domain is practicable, and can be of substantial value in industrial settings.

cs.AI↗

arXiv2020

Teacher-Student Framework Enhanced Multi-domain Dialogue Generation

Dialogue systems dealing with multi-domain tasks are highly required. How to record the state remains a key problem in a task-oriented dialogue system. Normally we use human-defined features as dialogue states and apply a state tracker to extract these features. However, the performance of such a system is limited by the error propagation of a state tracker. In this paper, we propose a dialogue generation model that needs no external state trackers and still benefits from human-labeled semantic data. By using a teacher-student framework, several teacher models are firstly trained in their individual domains, learn dialogue policies from labeled states. And then the learned knowledge and experience are merged and transferred to a universal student model, which takes raw utterance as its input. Experiments show that the dialogue system trained under our framework outperforms the one uses a belief tracker.

cs.CL↗

arXiv2020

Signal processing on simplicial complexes

Theoretical development and applications of graph signal processing (GSP) have attracted much attention. In classical GSP, the underlying structures are restricted in terms of dimensionality. A graph is a combinatorial object that models binary relations, and it does not directly model complex n-ary relations. One possible high dimensional generalization of graphs are simplicial complexes. They are a step between the constrained case of graphs and the general case of hypergraphs. In this paper, we develop a signal processing framework on simplicial complexes, such that we recover the traditional GSP theory when restricted to signals on graphs. It is worth mentioning that the framework works much more generally, though the focus of the paper is on simplicial complexes. We demonstrate how to perform signal processing with the framework using numerical examples.

eess.SP↗

arXiv2020

MTSS: Learn from Multiple Domain Teachers and Become a Multi-domain Dialogue Expert

How to build a high-quality multi-domain dialogue system is a challenging work due to its complicated and entangled dialogue state space among each domain, which seriously limits the quality of dialogue policy, and further affects the generated response. In this paper, we propose a novel method to acquire a satisfying policy and subtly circumvent the knotty dialogue state representation problem in the multi-domain setting. Inspired by real school teaching scenarios, our method is composed of multiple domain-specific teachers and a universal student. Each individual teacher only focuses on one specific domain and learns its corresponding domain knowledge and dialogue policy based on a precisely extracted single domain dialogue state representation. Then, these domain-specific teachers impart their domain knowledge and policies to a universal student model and collectively make this student model a multi-domain dialogue expert. Experiment results show that our method reaches competitive results with SOTAs in both multi-domain and single domain setting.

cs.CL↗

arXiv2020

GFCN: A New Graph Convolutional Network Based on Parallel Flows

In view of the huge success of convolution neural networks (CNN) for image classification and object recognition, there have been attempts to generalize the method to general graph-structured data. One major direction is based on spectral graph theory and graph signal processing. In this paper, we study the problem from a completely different perspective, by introducing parallel flow decomposition of graphs. The essential idea is to decompose a graph into families of non-intersecting one dimensional (1D) paths, after which, we may apply a 1D CNN along each family of paths. We demonstrate that the our method, which we call GraphFlow, is able to transfer CNN architectures to general graphs. To show the effectiveness of our approach, we test our method on the classical MNIST dataset, synthetic datasets on network information propagation and a news article classification dataset.

cs.LG↗

arXiv2020

Folded Graph Signals: Sensing with Unlimited Dynamic Range

Self-reset analog-to-digital converters (ADCs) are used to sample high dynamic range signals resulting in modulo-operation based folded signal samples. We consider the case where each vertex of a graph (e.g., sensors in a network) is equipped with a self-reset ADC and senses a time series. Graph sampling allows the graph time series to be represented by the signals at a subset of sampled vertices and time instances. We investigate the problem of recovering bandlimited continuous-time graph signals from folded signal samples. We derive sufficient conditions to achieve successful recovery of the graph signal from the folded signal samples, which can be achieved via integer programming. To resolve the scalability issue of integer programming, we propose a sparse optimization recovery method for graph signals satisfying certain technical conditions. Such an approach requires a novel graph sampling scheme that selects vertices with small signal variation. The proposed algorithm exploits the inherent relationship among the graph vertices in both the vertex and time domains to recover the graph signal. Simulations and experiments on images validate the feasibility of our proposed approach.

eess.SP↗

arXiv2019

Query-bag Matching with Mutual Coverage for Information-seeking Conversations in E-commerce

Information-seeking conversation system aims at satisfying the information needs of users through conversations. Text matching between a user query and a pre-collected question is an important part of the information-seeking conversation in E-commerce. In the practical scenario, a sort of questions always correspond to a same answer. Naturally, these questions can form a bag. Learning the matching between user query and bag directly may improve the conversation performance, denoted as query-bag matching. Inspired by such opinion, we propose a query-bag matching model which mainly utilizes the mutual coverage between query and bag and measures the degree of the content in the query mentioned by the bag, and vice verse. In addition, the learned bag representation in word level helps find the main points of a bag in a fine grade and promotes the query-bag matching performance. Experiments on two datasets show the effectiveness of our model.

cs.CL↗

arXiv2019

Transformation of Dense and Sparse Text Representations

Sparsity is regarded as a desirable property of representations, especially in terms of explanation. However, its usage has been limited due to the gap with dense representations. Most NLP research progresses in recent years are based on dense representations. Thus the desirable property of sparsity cannot be leveraged. Inspired by Fourier Transformation, in this paper, we propose a novel Semantic Transformation method to bridge the dense and sparse spaces, which can facilitate the NLP research to shift from dense space to sparse space or to jointly use both spaces. The key idea of the proposed approach is to use a Forward Transformation to transform dense representations to sparse representations. Then some useful operations in the sparse space can be performed over the sparse representations, and the sparse representations can be used directly to perform downstream tasks such as text classification and natural language inference. Then, a Backward Transformation can also be carried out to transform those processed sparse representations to dense representations. Experiments using classification tasks and natural language inference task show that the proposed Semantic Transformation is effective.

cs.CL↗

arXiv2019

Task-Oriented Conversation Generation Using Heterogeneous Memory Networks

How to incorporate external knowledge into a neural dialogue model is critically important for dialogue systems to behave like real humans. To handle this problem, memory networks are usually a great choice and a promising way. However, existing memory networks do not perform well when leveraging heterogeneous information from different sources. In this paper, we propose a novel and versatile external memory networks called Heterogeneous Memory Networks (HMNs), to simultaneously utilize user utterances, dialogue history and background knowledge tuples. In our method, historical sequential dialogues are encoded and stored into the context-aware memory enhanced by gating mechanism while grounding knowledge tuples are encoded and stored into the context-free memory. During decoding, the decoder augmented with HMNs recurrently selects each word in one response utterance from these two memories and a general vocabulary. Experimental results on multiple real-world datasets show that HMNs significantly outperform the state-of-the-art data-driven task-oriented dialogue models in most domains.

cs.CL↗

arXiv2019

A Hilbert Space Theory of Generalized Graph Signal Processing

Graph signal processing (GSP) has become an important tool in many areas such as image processing, networking learning and analysis of social network data. In this paper, we propose a broader framework that not only encompasses traditional GSP as a special case, but also includes a hybrid framework of graph and classical signal processing over a continuous domain. Our framework relies extensively on concepts and tools from functional analysis to generalize traditional GSP to graph signals in a separable Hilbert space with infinite dimensions. We develop a concept analogous to Fourier transform for generalized GSP and the theory of filtering and sampling such signals.

eess.SP↗

arXiv2019

Simple and Effective Text Matching with Richer Alignment Features

In this paper, we present a fast and strong neural approach for general purpose text matching applications. We explore what is sufficient to build a fast and well-performed text matching model and propose to keep three key features available for inter-sequence alignment: original point-wise features, previous aligned features, and contextual features while simplifying all the remaining components. We conduct experiments on four well-studied benchmark datasets across tasks of natural language inference, paraphrase identification and answer selection. The performance of our model is on par with the state-of-the-art on all datasets with much fewer parameters and the inference speed is at least 6 times faster compared with similarly performed ones.

cs.CL↗

arXiv2019

Review-Driven Answer Generation for Product-Related Questions in E-Commerce

The users often have many product-related questions before they make a purchase decision in E-commerce. However, it is often time-consuming to examine each user review to identify the desired information. In this paper, we propose a novel review-driven framework for answer generation for product-related questions in E-commerce, named RAGE. We develope RAGE on the basis of the multi-layer convolutional architecture to facilitate speed-up of answer generation with the parallel computation. For each question, RAGE first extracts the relevant review snippets from the reviews of the corresponding product. Then, we devise a mechanism to identify the relevant information from the noise-prone review snippets and incorporate this information to guide the answer generation. The experiments on two real-world E-Commerce datasets show that the proposed RAGE significantly outperforms the existing alternatives in producing more accurate and informative answers in natural language. Moreover, RAGE takes much less time for both model training and answer generation than the existing RNN based generation models.

cs.CL↗

arXiv2019

On the Properties of Gromov Matrices and their Applications in Network Inference

The spanning tree heuristic is a commonly adopted procedure in network inference and estimation. It allows one to generalize an inference method developed for trees, which is usually based on a statistically rigorous approach, to a heuristic procedure for general graphs by (usually randomly) choosing a spanning tree in the graph to apply the approach developed for trees. However, there are an intractable number of spanning trees in a dense graph. In this paper, we represent a weighted tree with a matrix, which we call a Gromov matrix. We propose a method that constructs a family of Gromov matrices using convex combinations, which can be used for inference and estimation instead of a randomly selected spanning tree. This procedure increases the size of the candidate set and hence enhances the performance of the classical spanning tree heuristic. On the other hand, our new scheme is based on simple algebraic constructions using matrices, and hence is still computationally tractable. We discuss some applications on network inference and estimation to demonstrate the usefulness of the proposed method.

eess.SP↗

arXiv2019

Estimating Infection Sources in Networks Using Partial Timestamps

We study the problem of identifying infection sources in a network based on the network topology, and a subset of infection timestamps. In the case of a single infection source in a tree network, we derive the maximum likelihood estimator of the source and the unknown diffusion parameters. We then introduce a new heuristic involving an optimization over a parametrized family of Gromov matrices to develop a single source estimation algorithm for general graphs. Compared with the breadth-first search tree heuristic commonly adopted in the literature, simulations demonstrate that our approach achieves better estimation accuracy than several other benchmark algorithms, even though these require more information like the diffusion parameters. We next develop a multiple sources estimation algorithm for general graphs, which first partitions the graph into source candidate clusters, and then applies our single source estimation algorithm to each cluster. We show that if the graph is a tree, then each source candidate cluster contains at least one source. Simulations using synthetic and real networks, and experiments using real-world data suggest that our proposed algorithms are able to estimate the true infection source(s) to within a small number of hops with a small portion of the infection timestamps being observed.

cs.SI↗

arXiv2019

Network Topology Inference Using Information Cascades with Limited Statistical Knowledge

We study the problem of inferring network topology from information cascades, in which the amount of time taken for information to diffuse across an edge in the network follows an unknown distribution. Unlike previous studies, which assume knowledge of these distributions, we only require that diffusion along different edges in the network be independent together with limited moment information (e.g., the means). We introduce the concept of a separating vertex set for a graph, which is a set of vertices in which for any two given distinct vertices of the graph, there exists a vertex whose distance to them are different. We show that a necessary condition for reconstructing a tree perfectly using distance information between pairs of vertices is given by the size of an observed separating vertex set. We then propose an algorithm to recover the tree structure using infection times, whose differences have means corresponding to the distance between two vertices. To improve the accuracy of our algorithm, we propose the concept of redundant vertices, which allows us to perform averaging to better estimate the distance between two vertices. Though the theory is developed mainly for tree networks, we demonstrate how the algorithm can be extended heuristically to general graphs. Simulations using synthetic and real networks, and experiments using real-world data suggest that our proposed algorithm performs better than some current state-of-the-art network reconstruction methods.

cs.SI↗

arXiv2018

Learning to Selectively Transfer: Reinforced Transfer Learning for Deep Text Matching

Deep text matching approaches have been widely studied for many applications including question answering and information retrieval systems. To deal with a domain that has insufficient labeled data, these approaches can be used in a Transfer Learning (TL) setting to leverage labeled data from a resource-rich source domain. To achieve better performance, source domain data selection is essential in this process to prevent the "negative transfer" problem. However, the emerging deep transfer models do not fit well with most existing data selection methods, because the data selection policy and the transfer learning model are not jointly trained, leading to sub-optimal training efficiency. In this paper, we propose a novel reinforced data selector to select high-quality source domain data to help the TL model. Specifically, the data selector "acts" on the source domain data to find a subset for optimization of the TL model, and the performance of the TL model can provide "rewards" in turn to update the selector. We build the reinforced data selector based on the actor-critic framework and integrate it to a DNN based transfer learning model, resulting in a Reinforced Transfer Learning (RTL) method. We perform a thorough experimental evaluation on two major tasks for text matching, namely, paraphrase identification and natural language inference. Experimental results show the proposed RTL can significantly improve the performance of the TL model. We further investigate different settings of states, rewards, and policy optimization methods to examine the robustness of our method. Last, we conduct a case study on the selected data and find our method is able to select source domain data whose Wasserstein distance is close to the target domain data. This is reasonable and intuitive as such source domain data can provide more transferability power to the model.

cs.IR↗