Information Extraction
Semantic Search
Knowledge Graphs
InferKnow is an automated Text Analysis and Knowledge Extraction Platform that is capable of identifying and extracting hidden knowledge from a collection of text documents.
Using Artificial Intelligence, Text Mining, Natural Language Processing and Machine Learning techniques along with domain-specific data models it automatically extracts information from unstructured text documents and converts them into structured entities and relations of a Knowledge Graph. The extracted knowledge is available for further analysis and decision-making through Dashboards, Semantic Search, Visualizations and queries to the Knowledge Graph.
With the growing volume of information each day it is becoming increasing difficult to manually review and analyse all information before making a decision. In addition to the exponential growth of information, the nature in which the said information are stored also adds to the problem. It is estimated that almost 80% of the data captured are stored in textual formats like Word or PDF documents, email messages, blog posts, research articles, etc. The information stored in such formats are not readily available to be queried upon making them unavailable to support decision-making. This unavailability can result in loss of some critical or novel information which could be invaluable for businesses.
Therefore, it is essential to have an automated system that can periodically process the available text data sources and extract the relevant information to assist in decision-making. InferKnow provides this automated system, with the ability to customize the extraction and analysis processes for different domains.
InferKnow can be employed in situations where you want to identify, extract and track specific pieces of information along with evidence of their mentions in the source document. This makes InferKnow suitable for scenarios requiring evidence based research and analysis, as below.
Supports automatic batch processing of multiple document types including: PDF, DOC, DOCX, CSV, JSON, JSONL, HTML, TXT
Uses Named Entity Recognition to extract domain specific entities like diseases, medications, apart from person names, places and organisations.
Uses Relation Extraction to extract different types of relations between the extracted entities and their attributes.
Use Semantic search to search of all mentions of an entity even if they are expressed using different lexical forms.
Use the generated Knowledge Graph to explore the associations between entities and discover novel and hidden relationships.
Use the Visualization tools for analysing the extracted information and get actionable insights that can assist you in decision-making.
