The task of text-to-SQL parsing, which aims at converting natural language questions into executable SQL queries, has garnered increasing attention in recent years. One of the major challenges in text-to-SQL parsing is domain generalization, i.e., how to generalize well to unseen databases. Recently, the pre-trained text-to-text transformer model, namely T5, though not specialized for text-to-SQL parsing, has achieved state-of-the-art performance on standard benchmarks targeting domain generalization. In this work, we explore ways to further augment the pre-trained T5 model with specialized components for text-to-SQL parsing. Such components are expected to introduce structural inductive bias into text-to-SQL parsers thus improving the model’s capacity on (potentially multi-hop) reasoning, which is critical for generating structure-rich SQLs. To this end, we propose a new architecture GRAPHIX-T5, a mixed model with the standard pre-trained transformer model augmented by specially-designed graph-aware layers. Extensive experiments and analysis demonstrate the effectiveness of GRAPHIX-T5 across four text-to-SQL benchmarks: SPIDER, SYN, REALISTIC and DK. GRAPHIX-T5 surpasses all other T5-based parsers with a significant margin, achieving new state-of-the-art performance. Notably, GRAPHIX-T5-large reaches performance superior to the original T5-large by 5.7% on exact match (EM) accuracy and 6.6% on execution accuracy (EX). This even outperforms the T5-3B by 1.2% on EM and 1.5% on EX

Communities are prevalent in social networks, knowledge graphs, and biological networks. Recently, the topic of community search (CS), extracting a dense subgraph containing a query vertex q from a graph, has received great attention. However, existing CS solutions are designed for undirected graphs, and overlook directions of edges which potentially lose useful information carried on directions. In many applications (e.g., Twitter), users' relationships are often modeled as directed graphs (e.g., if a user a follows another user b, then there is an edge from a to b). In this paper, we study the problem of CS on directed graph. Given a vertex q of a graph G, we aim to find a densely connected subgraph containing q from G, in which vertices have strong interactions and high similarities, by using the minimum in/out-degrees metric. We first develop a baseline algorithm based on the concept of D-core. We further propose three index structures and corresponding query algorithms. Our experimental results on seven real graphs show that our solutions are very effective and efficient. For example, on a graph with over 1 billion of edges, we only need around 40mins to index it and $1\sim 2$ sec to answer a query.

Ge Qu, Jinyang Li, Bowen Li, Bowen Qin, Nan Huo, Chenhao Ma, Reynold Cheng. Findings of the Association for Computational Linguistics ACL 2024. 2024.

Communities are prevalent in social networks, knowledge graphs, and biological networks. Recently, the topic of community search (CS) has received plenty of attention. The CS problem aims to look for a dense subgraph that contains a query vertex. Existing CS solutions do not consider the spatial extent of a community. They can yield communities whose locations of vertices span large areas. In applications that facilitate setting social events (e.g., finding conference attendees to join a dinner), it is important to find groups of people who are physically close to each other, so it is desirable to have a spatial-aware community (or SAC), whose vertices are close structurally and spatially. Given a graph G and a query vertex q, we develop an exact solution to find the SAC containing q, but it cannot scale to large datasets, so we design three approximation algorithms. We further study the problem of continuous SAC search on a "dynamic spatial graph", whose vertices' locations change with time, and propose three fast solutions. We evaluate the solutions on both real and synthetic datasets, and the results show that SACs are better than communities returned by existing solutions. Moreover, our approximation solutions perform accurately and efficiently.