Thomas Pajor

Route Planning in Transportation Networks

We survey recent advances in algorithms for route planning in transportation networks. For road networks, we show that one can compute driving directions in milliseconds or less even at continental scale. A variety of techniques provide different trade-offs between preprocessing effort, space requirements, and query time. Some algorithms can answer queries in a fraction of a microsecond, while others can deal efficiently with real-time traffic. Journey planning on public transportation systems, although conceptually similar, is a significantly harder problem due to its inherent time-dependent and multicriteria nature. Although exact algorithms are fast enough for interactive queries on metropolitan transit systems, dealing with continent-sized instances requires simplifications or heavy preprocessing. The multimodal route planning problem, which seeks journeys combining schedule-based transportation (buses, trains) with unrestricted modes (walking, driving), is even harder, relying on approximate solutions even for metropolitan inputs.

Customizable route planning

We present an algorithm to compute shortest paths on continental road networks with arbitrary metrics (cost functions). The approach supports turn costs, enables real-time queries, and can incorporate a new metric in a few seconds--fast enough to support real-time traffic updates and personalized optimization functions. The amount of metric-specific data is a small fraction of the graph itself, which allows us to maintain several metrics in memory simultaneously.

Sketch-based Influence Maximization and Computation: Scaling up with Guarantees

Propagation of contagion through networks is a fundamental process. It is used to model the spread of information, influence, or a viral infection. Diffusion patterns can be specified by a probabilistic model, such as Independent Cascade (IC), or captured by a set of representative traces. Basic computational problems in the study of diffusion are influence queries (determining the potency of a specified seed set of nodes) and Influence Maximization (identifying the most influential seed set of a given size). Answering each influence query involves many edge traversals, and does not scale when there are many queries on very large graphs. The gold standard for Influence Maximization is the greedy algorithm, which iteratively adds to the seed set a node maximizing the marginal gain in influence. Greedy has a guaranteed approximation ratio of at least (1-1/e) and actually produces a sequence of nodes, with each prefix having approximation guarantee with respect to the same-size optimum. Since Greedy does not scale well beyond a few million edges, for larger inputs one must currently use either heuristics or alternative algorithms designed for a pre-specified small seed set size. We develop a novel sketch-based design for influence computation. Our greedy Sketch-based Influence Maximization (SKIM) algorithm scales to graphs with billions of edges, with one to two orders of magnitude speedup over the best greedy methods. It still has a guaranteed approximation ratio, and in practice its quality nearly matches that of exact greedy. We also present influence oracles, which use linear-time preprocessing to generate a small sketch for each node, allowing the influence of any seed set to be quickly answered from the sketches of its nodes.

Intriguingly Simple and Fast Transit Routing

This paper studies the problem of computing optimal journeys in dynamic public transit networks. We introduce a novel algorithmic framework, called Connection Scan Algorithm (CSA), to compute journeys. It organizes data as a single array of connections, which it scans once per query. Despite its simplicity, our algorithm is very versatile. We use it to solve earliest arrival and multi-criteria profile queries. Moreover, we extend it to handle the minimum expected arrival time (MEAT) problem, which incorporates stochastic delays on the vehicles and asks for a set of (alternative) journeys that in its entirety minimizes the user’s expected arrival time at the destination. Our experiments on the dense metropolitan network of London show that CSA computes MEAT queries, our most complex scenario, in 272 ms on average.

Customizable Route Planning in Road Networks

We propose the first routing engine for computing driving directions in large-scale road networks that satisfies all requirements of a real-world production system. It supports arbitrary metrics (cost functions) and turn costs, enables real-time queries, and can incorporate a new metric in less than a second, which is fast enough to support real-time traffic updates and personalized cost functions. The amount of metric-specific data is a small fraction of the graph itself, which allows us to maintain several metrics in memory simultaneously. The algorithm is the core of the routing engine currently in use by Bing Maps.

Accelerating Multi-modal Route Planning by Access-Nodes

Recent research on fast route planning algorithms focused either on road networks or on public transportation. However, on the long run, we are interested in planning routes in a multi-modal scenario: we start by car to reach the nearest train station, ride the train to the airport, fly to an airport near our destination and finally take a taxi. In other words, we need to incorporate public transportation into road networks. However, we do not want to switch the type of transportation too often. We end up in a label constrained variant of the shortest path problem. In this work, we present a first efficient solution to a restricted variant of this problem including experimental results for transportation networks with up to 125 Mio. edges.

Energy-optimal routes for electric vehicles

We study the problem of electric vehicle route planning, where an important aspect is computing paths that minimize energy consumption. Thereby, any method must cope with specific properties, such as recuperation, battery constraints (over- and under-charging), and frequently changing cost functions (e. g., due to weather conditions). This work presents a practical algorithm that quickly computes energy-optimal routes for networks of continental scale. Exploiting multi-level overlay graphs [25, 30], we extend the Customizable Route Planning approach [7] to our scenario in a sound manner. This includes the efficient computation of profile queries and the adaption of bidirectional search to battery constraints. Our experimental study uses detailed consumption data measured from a production vehicle (Peugeot iOn). It reveals for the network of Europe that a new cost function can be incorporated in about five seconds, after which we answer random queries within 0.3 ms on average. Additional evaluation on an artificial but realistic [21, 35] vehicle model with unlimited range demonstrates the excellent scalability of our algorithm: Even for long-range queries across Europe it achieves query times below 5 ms on average---fast enough for interactive applications. Altogether, our algorithm exhibits faster query times than previous approaches, while improving (metric-dependent) preprocessing time by three orders of magnitude.

Robust Distance Queries on Massive Networks

We present a versatile and scalable algorithm for computing exact distances on real-world networks with tens of millions of arcs in real time. Unlike existing approaches, preprocessing and queries are practical on a wide variety of inputs, such as social, communication, sensor, and road networks. We achieve this by providing a unified approach based on the concept of 2-hop labels, improving upon existing methods. In particular, we introduce a fast sampling-based algorithm to order vertices by importance, as well as effective compression techniques.

Engineering Time-Expanded Graphs for Faster Timetable Information

We present an extension of the well-known time-expanded approach for timetable information. By remodeling unimportant stations, we are able to obtain faster query times with less space consumption than the original model. Moreover, we show that our extensions harmonize well with speed-up techniques whose adaption to timetable networks is more challenging than one might expect.

Computing classic closeness centrality, at scale

Closeness centrality, first considered by Bavelas (1948), is an importance measure of a node in a network which is based on the distances from the node to all other nodes. The classic definition, proposed by Bavelas (1950), Beauchamp (1965), and Sabidussi (1966), is (the inverse of) the average distance to all other nodes. We propose the first highly scalable (near linear-time processing and linear space overhead) algorithm for estimating, within a small relative error, the classic closeness centralities of all nodes in the graph. Our algorithm applies to undirected graphs, as well as for centrality computed with respect to round-trip distances in directed graphs. For directed graphs, we also propose an efficient algorithm that approximates generalizations of classic closeness centrality to outbound and inbound centralities. Although it does not provide worst-case theoretical approximation guarantees, it is designed to perform well on real networks. We perform extensive experiments on large networks, demonstrating high scalability and accuracy.