“deecoob insight” – Software-as-a-Service for text mining, machine learning and visualisation of data

deecoob insight is powerful, scalable and performant. The platform is able to crawl large amounts of data, automatically check their relevance, compare data with customer contacts, license or contract databases as well as other predefined sources and properties. deecoob insight is implemented on modern, state of the art, intelligent search technologies (e.g. Elasticsearch). These index, analyze and verify the data collected from the sources. Complex algorithms and data mining processes as well as methods of artificial intelligence, linguistics, semantics, statistics, forensics and heuristic procedures classify and cluster the mass data meaningfully and provide them with rankings and scores.

Data mining

The processing stages of deecoob insight are characterized by complex data mining processes. With the help of semi-supervised learning algorithms for automated text analysis and text classification (e.g. naive bayes) and state-of-the-art technologies for text mining (e.g. support vector machines, neural networks), deecoob insight is able to process the collected structured and unstructured mass data efficiently for customer requirements. Thus, several million records are iteratively, multi-level and continuously compressed to a few thousand or hundreds of actual data relevant for the custumer.

Machine learning

Parallel to data mining for filtering and compression of data on deecoob insight, permanent feedback loops run on the system thus enabling self optimizing procedures (machine learning). In many cases, certain properties (e.g. relevant, primary, complete, current) are automatically assigned to the data from the software. The research professionals then determine the correct properties relevant to the data during manual qualification. These properties can be “not relevant, secondary information, incomplete, obsolete”. deecoob insight´s self-learning algorithms determine this difference (feedback loop) and automatically adjust for future evaluations. In terms of artificial intelligence, deecoob insight is therefore a self-learning and self-optimizing software.

Active learning

Further feedback loops run between customers, research professionals (e.g. deecoob insight users) and the platform. Customers normally choose to work on a long term basis with deecoob insight and deecoob service. This gives all participants the opportunity to continuously carry out measurements, monitorings, retrospective and iterative evaluations of the quality and quantity of results as well as customer satisfaction. Actions for optimizing and expanding the platform are derived from this. In this case, these actions are not implemented by self-learning forces (see machine learning) but thru deecoob´s data scientists and software engineers. deecoob insight is then able to crawl additional or different data to answer new or different customer questions and satisfy new or additional customer requirements.

Data-as-a-service or software-as-a-service

As a customer, you can use deecoob insight as data-as-a-service (DaaS) or software-as-a-service (SaaS). The software infrastructure is always operated by deecoob, so you do not have to invest in server hardware or server software.

If you choose the DaaS model, deecoob will provide you the exact data and documents you require. deecoob does the job for you. Apart from service and project costs, you do not have any further expenses for hardware, software, access points or employees.

If you prefer the SaaS model, your employees work as research professionals with the deecoob insight client. The accesses to deecoob insight (e.g. VPN, virtual environment) will be provided and managed by deecoob. Your employees need be trained to use deecoob insight. The hardware and software for this must be provided by you.

Visualization & monitoring

deecoob insight constantly runs extensive logging processes in the background. All data, processes and events are recorded and stored. The decisive factor for managing the platform and quality of data and processes, is to operate an intelligent monitoring. With Kibana, the software Elasticsearch provides a modern product for visualization of mass data and provision of dashboards.

You define individual KPI (key performance indicators) with our data scientists. We implement the necessary dashboards for your permanent monitoring and measuring of your success. It does not matter whether you wish to determine the output of a specific source, the productivity of users, the composition of data, the geographical or temporal distribution of data or other facts. Kibana gives you the right answer in the form of bar, line or pie charts, heatmaps or in many other formats. Standard functions such as drill-down, drill-out, discover, dashboard sharing, etc. are main features of Kibana.

Software architecture

deecoob implements and operates the software platform deecoob insight. In doing so, we mainly rely on open source products that have already proven themselves in similar complex and sophisticated software architectures. Our claim is the continuous availability of deecoob insight and uninterrupted procurement and qualification of data for our customers. Therefore, we are not satisfied with compromises.

Customers expect and receive a modern, performant, scalable, configurable and expandable platform. Standardized interfaces allow the integration of additional sources for data crawling or the connection of customer systems for data transfer in a short time and with little effort. The platform simultaneously supplies several customer demands. Nevertheless, each customer’s data, information, models and workflows are protected and separated from each other.

Test deecoob insight now.