Graphs representing our results for different topics and metrics.

Graph data represents complex relationships across diverse domains, from social networks to healthcare and chemical sciences. However, real-world graph data often spans multiple modalities, including time-varying signals from sensors, semantic information from textual representations, and domain-specific encodings. This dissertation introduces innovative multimodal learning techniques for graph-based predictive modeling, addressing the intricate nature of these multidimensional data representations. The research systematically advances graph learning through innovative methodological approaches across three critical modalities. Initially, we establish robust graph-based methodological foundations through advanced techniques including prompt tuning for heterogeneous graphs and a comprehensive framework for imbalanced learning on graph data. we then extend these methods to time series analysis, demonstrating their practical utility through applications such as hierarchical spatio-temporal modeling for COVID-19 forecasting and graph-based density estimation for anomaly detection in unmanned aerial systems. Finally, we explore textual representations of graphs in the chemical domain, reformulating reaction yield prediction as an imbalanced regression problem to enhance performance in underrepresented high-yield regions critical to chemists.

This data supports the assertions made in a paper (same name as this project) which surveyed 3 Second Language Acquisition journals (Modern Language Journal, Language Learning, and Studies in Second Language Acquisition) from the time of their inception to 2011/2012. The raw data used for calculations about the number of graphics and which type of graphics were published is included in the attached Excel file.

The 16th OpenAIRE Graph Community Call took place on Wednesday 18 June 2025, where Andrea Mannocci, OpenAIRE Graph Data Scientist, presented the Scholarly/Scientific Knowledge Graph Interoperability Framework (SKG-IF). This presentation is part of the Community Call series where the OpenAIRE Graph team dives into the makings and workings of the OpenAIRE Graph, one of the world’s largest Scholarly Knowledge Graphs, and give you the floor for questions, feedback, & suggestions. You can view and register for upcoming calls and consult all past call materials on the OpenAIRE Graph website. Recording: https://youtu.be/Fai8BwEZC6w

Yearly citation counts for the publication titled "Polynomial‐time algorithms for solving a class of critical node problems on trees and series‐parallel graphs".

A critical issue in intelligent building control is detecting energy consumption anomalies based on intelligent device status data. The building field is plagued by energy consumption anomalies caused by a number of factors, many of which are associated with one another in apparent temporal relationships. For the detection of abnormalities, most traditional detection methods rely solely on a single variable of energy consumption data and its time series changes. Therefore, they are unable to examine the correlation between the multiple characteristic factors that affect energy consumption anomalies and their relationship in time. The outcomes of anomaly detection are one-sided. To address the above problems, this paper proposes an anomaly detection method based on multivariate time series. Firstly, in order to extract the correlation between different feature variables affecting energy consumption, this paper introduces a graph convolutional network to build an anomaly detection framework. Secondly, as different feature variables have different influences on each other, the framework is enhanced by a graph attention mechanism so that time series features with higher influence on energy consumption are given more attention weights, resulting in better anomaly detection of building energy consumption. Finally, the effectiveness of this paper’s method and existing methods for detecting energy consumption anomalies in smart buildings are compared using standard data sets. The experimental results show that the model has better detection accuracy.

The global temperature time series provides time series charts using station based observations of daily temperature. These charts provide information about the observations compared to the derived daily normal temperature for various time scales (30, 90, 365 days). Each station has a graphic that contains three charts. The first chart in the graphic is a time series in the format of a line graph, representing the daily average temperatures compared to the expected daily normal temperatures. The second chart is a bar graph displaying daily departures from normal, including a line depicting the mean departure for the period. The third chart is a time series of the observed daily maximum and minimum temperatures. The graphics are updated daily and the graphics reflect the updated observations including the latest daily data available. The available graphics are rotated, meaning that only the most recently created graphics are available. Previously made graphics are not archived.

Abstract

The dataset was derived by the Bioregional Assessment Programme from multiple source datasets. The source datasets are identified in the Lineage field in this metadata statement. The processes undertaken to produce this derived dataset are described in the History field in this metadata statement.

This dataset contains time series figures (shown in the report) generated for baseline and crdp mine footprints , which represent the footprints used in the surface water modelling. The footprints are contained within a single shapefile (HUN Mine footprints for timeseries) and the timelines contained within the the spreadhseet (HUN mine time series tables v01).

Dataset History

The footprints are contained within a single shapefile (HUN Mine footprints for timeseries) and the timelines contained within the the spreadsheet (HUN mine time series tables v01). Timelines for all mines were assembled into the spreadsheet Mine_files_summary_Final.xlsx. The script MineFootprint_TimeSeries_Final.m reads the data from the spreadsheet and creates the time series figures in png format which form the dataset.

Dataset Citation

Bioregional Assessment Programme (XXXX) HUN Mine Footprints Timeseries Graph v01. Bioregional Assessment Derived Dataset. Viewed 22 June 2018, http://data.bioregionalassessments.gov.au/dataset/11493517-df5f-49ed-84dc-23afdbe00c5e.

Dataset Ancestors

• Derived From HUN Groundwater footprint polygons v01

• Derived From HUN mine time series tables v01

• Derived From BILO Gridded Climate Data: Daily Climate Data for each year from 1900 to 2012

• Derived From HUN Historical Landsat Images Mine Foot Prints v01

• Derived From Historical Mining footprints DTIRIS HUN 20150707

• Derived From HUN Mine footprints for timeseries

• Derived From Climate model 0.05x0.05 cells and cell centroids

• Derived From HUN Historical Landsat Derived Mine Foot Prints v01

• Derived From HUN SW footprint shapefiles v01

• Derived From Mean Annual Climate Data of Australia 1981 to 2012

Zhang et al. (https://link.springer.com/article/10.1140/epjb/e2017-80122-8) suggest a temporal random network with changing dynamics that follow a Markov process, allowing for a continuous-time network history moving from a static definition of a random graph with a fixed number of nodes n and edge probability p to a temporal one. Defining lambda = probability per time granule of a new edge to appear and mu = probability per time granule of an existing edge to disappear, Zhang et al. show that the equilibrium probability of an edge is p=lambda/(lambda+mu) Our implementation, a Python package that we refer to as RandomDynamicGraph https://github.com/ScanLab-ossi/DynamicRandomGraphs, generates large-scale dynamic random graphs according to the defined density. The package focuses on massive data generation; it uses efficient math calculations, writes to file instead of in-memory when datasets are too large, and supports multi-processing. Please note the datetime is arbitrary.

MS-BioGraphs are a family of sequence similarity graph datasets with up to 2.5 trillion edges. The graphs are weighted edges and presented in compressed WebGraph format. The dataset include symmetric and asymmetric graphs. The largest graph has been created by matching sequences in Metaclust dataset with 1.7 billion sequences. These real-world graph dataset are useful for measuring contributions in High-Performance Computing and High-Performance Graph Processing.

Description. This is the data used in the experiment of the following conference paper:

• N. Arınık, R. Figueiredo, and V. Labatut, “Signed Graph Analysis for the Interpretation of Voting Behavior,” in International Conference on Knowledge Technologies and Data-driven Business - International Workshop on Social Network Analysis and Digital Humanities, Graz, AT, 2017, vol. 2025. ⟨hal-01583133⟩

Source code. The code source is accessible on GitHub: https://github.com/CompNet/NetVotes

Citation. If you use the data or source code, please cite the above paper.

@InProceedings{Arinik2017,
author = {Arınık, Nejat and Figueiredo, Rosa and Labatut, Vincent},
title = {Signed Graph Analysis for the Interpretation of Voting Behavior},
booktitle = {International Conference on Knowledge Technologies and Data-driven Business - International Workshop on Social Network Analysis and Digital Humanities},
year = {2017},
volume = {2025},
series = {CEUR Workshop Proceedings},
address = {Graz, AT},
url = {http://ceur-ws.org/Vol-2025/paper_rssna_1.pdf},
}

----------------------

Details.

# RAW INPUT FILES
The 'itsyourparliament' folder contains all raw input files for further data processing (such as network extraction).
The folder structure is as follows:
* itsyourparliament/
** domains: There are 28 domain files. Each file corresponds to a domain (such as Agriculture, Economy, etc.) and contains corresponding vote identifiers and their "itsyourparliament.eu" links.
** meps: There are 870 Member of Parliament (MEP) files. Each file contains the MEP information (such as name, country, address, etc.)
** votes: There are 7513 vote files. Each file contains the votes expressed by MEPs
# NETWORKS AND CORRESPONDING PARTITIONS
This work studies the voting behavior of French and Italian MEPs on "Agriculture and Rural Development" (AGRI) and "Economic and Monetary Affairs" (ECON) for each separate year of the 7th EP term (2009-10, 2010-11, 2011-12, 2012-13, 2013-14). Note that the interpretation part (section 4) of the published paper is limited to only a few of these instances (2009-10 in ECON and 2012-13 in AGRI).
The extracted networks are located in the "networks" folder and the corresponding partitions are in the "partitions" folder. Both folders have the same structure, which is as follows:
COUNTRY-NAME
|_DOMAIN-NAME
|_2009-10
|_2010-11
|_2011-12
|_2012-13
|_2013-14
## NETWORKS
The networks in this folder are used in the article. All those networks are the ones obtained after the filtering step (as explained in the article). The networks are in 'Graphml' format. These networks are enriched with some MEPs' properties (such as name, political party, etc.) associated with each node.
## ALL NETWORKS
For those who are interested in other countries or domains, we make available all possible networks that we can extract from raw data with vs. without filtering step.
COUNTRY-NAME
|_m3
|_negtr=NA_postr=NA: This folder contains all filtered networks. Note that the filtering step is explained in Section 2.1.2 of the article.
|_bygroup
|_bycountry
|_negtr=0_postr=0: This folder contains all original networks (i.e. no filtering step).
|_bygroup
|_bycountry
## PARTITIONS
The partitions are obtained in this way: First, the Ex-CC (exact) method is run and we denote 'k' for the the number of detected cluster in output. This 'k' value is the reference point in order to run the ILS-RCC (heuristic) method by specifying the number of desired cluster in output. Then, ILS-RCC is run with various values ('k', 'k+1', 'k+2'). All those results are integrated into the initial network graphml files and then converted into gephi format so that this will help dive in the results in interactive way.
Note that we need to handle the absent MEPs in clustering results. Because, those MEPs correspond to isolated nodes in networks. Each isolated node is considered a single cluster node in Ex-CC results. We simply omit those nodes in order to find the 'k' (number of detected cluster) value before running ILS-RCC. Not also that ILS-RCC does not process isolated nodes such that an isolated node can be part of a cluster.

----------------------
# COMPARISON RESULTS
The 'material-stats' folder contains all the comparison results obtained for Ex-CC and ILS-CC. The csv files associated with plots are also provided.
The folder structure is as follows:
* material-stats/
** execTimePerf: The plot shows the execution time of Ex-CC and ILS-CC based on randomly generated complete networks of different size.
** graphStructureAnalysis: The plots show the weights and links statistics for all instances.
** ILS-CC-vs-Ex-CC: The folder contains 4 different comparisons between Ex-CC and ILS-CC: Imbalance difference, number of detected clusters, difference of the number of detected clusters, NMI (Normalized Mutual Information)

----------------------
Funding: Agorantic FR 3621, FMJH Program Gaspard Monge in optimization and operation research (Project 2015-2842H)

Graph and download economic data for Population, Total for Cameroon (POPTOTCMA647NWDB) from 1960 to 2024 about Cameroon and population.

Random Graphs

  Dataset Description

Over 20 000 randomly generated directed graphs with labels and formatting assigned at random.

  Key Features

Generation method: Erdős–Rényi random graph model
Vertices: 2 to 15 per graph
Edge probability: Uniformly random between 0.0 and 1.0
Labels: Sampled from the agentlans/noun-phrases dataset

  Dataset Structure

Each entry includes:

image: PNG rendering of the graph
dot: Graphviz DOT source code
lisp:… See the full description on the dataset page: https://huggingface.co/datasets/agentlans/random-graphs.

The graph technology market is experiencing robust growth, driven by the increasing need for advanced data analytics and the rising adoption of artificial intelligence (AI) and machine learning (ML) applications. The market's expansion is fueled by the ability of graph databases to handle complex, interconnected data more efficiently than traditional relational databases. This is particularly crucial in industries like finance (fraud detection, risk management), healthcare (patient relationship mapping, drug discovery), and e-commerce (recommendation systems, personalized marketing). Key trends include the move towards cloud-based graph solutions, the integration of graph technology with other data management systems, and the development of more sophisticated graph algorithms for advanced analytics. While challenges remain, such as the need for skilled professionals and the complexity of implementing graph databases, the overall market outlook remains positive, with a projected Compound Annual Growth Rate (CAGR) – let's conservatively estimate this at 25% – for the forecast period 2025-2033. This growth will be driven by ongoing digital transformation initiatives across various sectors, leading to an increased demand for efficient data management and analytics capabilities. We can expect to see continued innovation in both open-source and commercial graph database solutions, further fueling the market's expansion. The competitive landscape is characterized by a mix of established players like Oracle, IBM, and Microsoft, alongside emerging innovative companies such as Neo4j, TigerGraph, and Amazon Web Services. These companies are constantly vying for market share through product innovation, strategic partnerships, and acquisitions. The presence of both open-source and proprietary solutions caters to a diverse range of needs and budgets. The market segmentation, while not explicitly detailed, likely includes categories based on deployment (cloud, on-premise), database type (property graph, RDF), and industry vertical. The regional distribution will likely show strong growth in North America and Europe, reflecting the higher adoption of advanced technologies in these regions, followed by a steady rise in Asia-Pacific and other developing markets. Looking ahead, the convergence of graph technology with other emerging technologies like blockchain and the Internet of Things (IoT) promises to unlock even greater opportunities for growth and innovation in the years to come.

Leading Indicators OECD: Component series: Orders: Original series for the United States was 295357000000.00000 US $ in November of 2023, according to the United States Federal Reserve. Historically, Leading Indicators OECD: Component series: Orders: Original series for the United States reached a record high of 301262000000.00000 in June of 2023 and a record low of 14092000000.00000 in January of 1961. Trading Economics provides the current actual value, an historical data chart and related indicators for Leading Indicators OECD: Component series: Orders: Original series for the United States - last updated from the United States Federal Reserve on September of 2025.

This dataset provides reference ontologies that were translated from product design and inspection data from the National Institute of Standards and Technology (NIST) Smart Manufacturing Systems (SMS) Test Bed. The examples represents a three-component assembly of a box, machined from Aluminum, and has a technical data package available on the SMS Test Bed website. The use of the ontologies aims to integrate the product lifecycle data of engineering design represented in the STEP AP242 format, which is described in the ISO 10303 series, as well as quality assurance data, representing in the Quality Information Framework (QIF) standard.

Index Time Series for Global X NASDAQ 100® Risk Managed Income ETF. The frequency of the observation is daily. Moving average series are also typically included. The fund invests at least 80% of its total assets in the securities of the Nasdaq-100 Monthly Net Credit Collar 95-100 Index (underlying index). The underlying index measures the performance of a risk managed income strategy that holds the underlying stocks of the NASDAQ 100® Index and applies an options collar strategy (i.e., a mix of short (sold) call options and long (purchased) put options) on the NASDAQ 100® Index. The fund is non-diversified.

Index Time Series for SPDR® S&P 600 Small Cap Growth ETF. The frequency of the observation is daily. Moving average series are also typically included. The fund generally invests substantially all, but at least 80%, of its total assets in the securities comprising the index. The index measures the performance of the small-capitalization growth segment of the U.S. equity market. It may purchase a subset of the securities in the index in an effort to hold a portfolio of securities with generally the same risk and return characteristics of the index.

Leading Indicators OECD: Component series: Share prices: Normalised for the United States was 99.78373 Index in December of 2023, according to the United States Federal Reserve. Historically, Leading Indicators OECD: Component series: Share prices: Normalised for the United States reached a record high of 103.83848 in August of 2007 and a record low of 95.60843 in March of 2009. Trading Economics provides the current actual value, an historical data chart and related indicators for Leading Indicators OECD: Component series: Share prices: Normalised for the United States - last updated from the United States Federal Reserve on August of 2025.

This dataset contains graphs and a movie. Both show visualizations of rotational curves in the transversal plane within a Standardized Gait Cycle from Vertebra prominens downwards, ending at the pelvis. They display 201 anonymous healthy people aged 18-70 years walking at 2,3,4, and 5 km/h on a treadmill. They are based on a SPSS (v23) syntax file and a relating graph template that can be found at our datasets as well. Files are numbered subsequently across all speeds and can be linked by number to its non-standardized counterpart in a further dataset. Positive values show vertebral body rotation to the left, negative values show rotation to the right. Percent of the Standardized Gait Cycle (0-100%) is displayed on the abscissa, always starting with Initial Contact of the right foot. Within a Standardized Gait Cycle the duration of the stance phase right is expected to be 60% (Perry, 1992). As can be seen in the graphs, interpolating spline functions work for average walking speed measurements leading to a more precise determination of relevant and characteristic points (e.g. maxima, phase shifts, lumbar and thoracic movement behavior), thereby aiding in in the clarification of individual features.