Respond Projects

Research Areas

Algorithms for knowledge extraction from big data

Large volumes of data that cannot be stored in normal relational databases are being generated every day from the remote sensing satellites. Many software elements extract information from the raw data generating information in unstructured form such as images, log files, user orders in pdf, word etc. There is a need for developing efficient data mining algorithms to tag the data sets for facilitating efficient build up of archival and retrieval. In general data mining algorithms work on data sets that are of reasonable size and cannot handle BIG data. Develop Parallel Algorithms for mining the classification rules to facilitate data archival in an optimal manner Develop mining algorithms that are Incremental and can learn and unlearn from the continuous satellite data acquisitions Develop algorithms for extracting meaningful trends in the customer ordering, build customer satisfaction index, predict the future sales or potential sensors or popular products etc

Software reliability modelling and metrics

There is a need to develop automated tools to extract different metrics from various software packages developed by ISRO to estimate their reliability and predict if possible the failure rates from the version history. Develop customized metrics for different types of software packages including real time, near real time, post processing, workflow software and distributed software Develop algorithms for estimating the software reliability numbers and predictive models for forecasting the failure conditions.