BIG Data, data analytics problems addressed at DISIT

In this page, some of the big data, data analytics problems at DISIT are listed referring to the projects in which they are addressed:

• Artificial Intelligence, Machine learning, Explainable AI, XAI
  • https://www.snap4city.org/download/video/course2020/da/
  • techniques
    • RF, XGBoost, SVM, BRNNm SVR, MLP, ....
    • LSTM, BI-LSST, CONV-LSTM, etc., 
    • Clustering: Kmeans, K-Meodis, hierarchical clustering, etc. 
    • What-IF Analysis….
  • Models and solutions 
    • predicting NO2 from traffic flow
    • predicting free parking lots
    • predicting free bike sharing lots, bike sharing returning slots
    • predicting traffic flow data
    • predicting NO2 long term
    • anomaly detection on traffic flow
    • anomaly detection from IOT data
    • Smart parking: Predictions
    • User Behavior Analysis, via  Wi-Fi, OD, trajectories
    • Recognition of Used Transportation means
    • Traffic Flow Reconstruction, from Traffic Sensors Data
    • Covid-19 vs other data: traffic and environmental
    • Quality of Public Transport Service
    • Origin Destination Matrices from: Wi-Fi, Mobile Apps, etc.
    • Demand of Mobility vs Offer of Transportation
    • Modal and Multimodal Routing for  Navigation and Travel Planning
    • Environmental Data Analysis and Predictions, early Warning
    • Prediction of Air Quality Conditions
    • 15MinCityIndex Assessment of Cities
    • Predictive Maintenance on Plant
    • Fashion Retail shop recommendations
  •  
• Semantic Computing:
  • ontological modeling and tools: smart city, smart cloud, brain, railways, video annotation, cultural heritage, etc. etc.
  • brain modeling and ontology
  • Ontological models, ontology engineering
  • Ontology and semantic model for Cloud, smart cloud algorithms
  • Ontology and semantic model interlocking and signalling, train and driverless metro
• Data Mining and Data Fusion:
  • data mining for heterogeneous data, data reconciliation, data fusion
  • data mining and data warehouse for smart city data (open data, from public and private), see Smart City projects as: sii-mobility project
  • data mining of text in English and Italian, affective analysis,
  • knowledge modeling and understanding, knowledge indexing, knowledge indexing and search; see projects as OSIM, and SACVAR projects when all the Tuscany web pages are crawled and processed
  • Smart City processes, tools and services http://www.disit.org/6566
  • Geolocation extractor: http://www.disit.org/6540
  • Blog Vigilance and Twitter vigilance via Natural Language Processing
  • data quality improvement
  • DIM: Data Ingestion Manager: http://www.disit.org/6732
• LOD, RDF Stores, Graph Database:
  • linked data management, distributed query and rendering, LD and LOD, RDF stores integration and queries, see LOG.DISIT.ORG service.
  • performance analysis of RDF stores
  • Indexer Manager for shortenining time in RDF indexing and versioniong
  • RDF store and dbGraph methodology and tools
• Data Analytics:
  • Fault prediction: logistic regressions, regressions, etc.
  • prediction of people flow in the city
  • prediction of traffic flow status in the city
  • prediction of audience on tv shows
  • understanding user behaviour in the city, and on mobile devices
  • smart decision support system, http://smartds.disit.org, via computing and system thinking
  • Data indexing and search, cross media indexing: taboo, genetic algorithms, etc.
  • Keyword extraction, data mining extraction
  • recommendations by similarity and by complementarity (matchmaking)
  • clustering, hierarchical and incremental clustering, k-means, k-medoids
  • link discovering and reconciliation, single and multiple RDF Stores
  • computing origin destination matrix, prediction people flow
  • identification of optimum Access Point WiFi position in the city
  • Twitter vigilance via Natural Language Processing
• Architectures and system solutions
  • Natural Language Processing on Hadoop
  • A Distributed Framework for NLP-Based Keyword and Keyphrase Extraction From Web Pages and Documents, DMS 2015 best paper award on NLP for Hadoop
  • parallel and distributed processing via media grid, for smart cloud e smart city.
  • P2P architectures and solution for DRM indexing and distributed licensing
  • massive fully distributed scheduling for data gathering and processing, SCE
  • Blog and Twitter vigilance
  • RDF store assessment, comparison among dbGraph, Virtuoso, etc.
  • Benchmark for RDF store assessment
• Applications:
  • See Km4City page
  • See Km4City demonstrators and real tools
  • mainly on: smart city, smart cloud, cultural heritage, P2P DRM, railways, transportation, ...
  • Big Data Smart City processes and tools, Dec 2014
  • Smart City DISIT page and km4city.
  • identification of critical conditions and predictions, see projects as: sii-mobility for smart city, and ICARO for smart cloud
  • cross media and multi language indexing and search, semantic indexing: see ECLAP project and service
  • conditional access and digital rights management, managing grant authorization for hundreds thousands of users at the same time instant, managing licenses. See project AXMEDIS and the DISIT extension on P2P grant authorization and licensing.
  • keyword extraction on massive number of documents per seconds: see ECLAP and SACVAR projects and tools.
  • html web pages mining for geolocation of them with high precision: see sii-mobility project.
  • content based information retrieval of images and text / metadata, and integration, see XMediaCBIR.
• Courses:
  • Data ingestion and management
  • smart city solutions
  • Hadoop architecture and programming
  • comparison of Big Data noSQL databases
  • Semantic computing, knowledge and ontology engineering (many)
  • suggestion and recommendations
  • anatomy of a social network

 
Most of these solutions are working on scalable cloud solutions based on grid parallel processing and/on on Hadoop.
Last Update: 2017/01/15