The project uses a High-performance cluster composed of a precise data extractor, a hypercube and a search engine capable of indexing the whole web in a immediate-like time and presents information in a structured format. The project uses also a high-level Natural language processing tools that do the segmentation and the linguistic analysis based on human morpho-syntactic inference grammars to extract and classify information in terms of domain, importance and sentiment, this makes the solution also quickly portable to multiple languages with low engineering cost and high precision.
The following example shows graphical statistics showing the amount of information published in websites for the financial domain of Apple's news.
This example shows a query per scenario of the financial news of Apple's company. Multiple queries can be made for other legal, economic, product reviews, technology ... and unlimited domain scenarios. Quereis are quickly adapted for client needs and are saved to provide periodical growth reporting.
This is a real scientific example made for knowledge extraction of regions and their locations, each results is an extracted named entity connected to its generic type. It identifies "Emarati popular club, Castle of persia, Quansoun ..." as locations.
The next graphics are diagrams published with Excel of Street furniture market, the data has been extracted by our precise data extraction tool. The study has been made by a product manager and our.