Analytics & Visualization

Table of Contents

Visualizing Complex Data

Our analytical and visualization tools help data scientists visualize and decipher complex data in an intuitive manner. The tools are scalable & can be deployed as enterprise analytical and delivery platforms.

Parallels

Explore

  • Start with preferences
  • Explore via:
    • Topics
    • Content types
    • Concepts
    • Content metadata

Analyse

  • Top actors in the notifications pool
  • Most common interactions amongst actors

Spot trends

  • Profiling
  • Generic trends
  • Gap analysis

Case Study

Problem Statement

  • Solution for a Personalized research content alert and analytics system

  • Given a pool of diverse content, both ready to consume and to be acquired from open-source periodically, how do we alert users  on the most relevant  content in the subject(s)/area(s) of their choosing? How can we make sure we cover most of the latest research?

  • How do we make sure we cater to specific user preferences without explicit input(s) or overpowering users with heavy volumes?

  • We need a solution that is sustainable; robust in terms of core capabilities and flexible in terms of content and functionalities

Challenges

Users are allowed to enter almost free text selection of keywords/phrases with minimal syntax.Onus is on the solution envisioned to incorporate a query parsing component

Multidimensional content and metadata to be leveraged across sources with due diligence to update cycles

Completely automated preference tracker and a feedback inclusion system to act on and update the system to tailor the alerts to the preference pattern with immediate effect

A adaptive workflow system to ingest incoming user subscriptions, unsubscribe  requests, mail tracker and change management

Analytics and decision aids that point the decision makers towards how well their content is being accepted, consumed and used. This system ties in content, user behaviour and approaches used for suggesting relevant alerts

Solutions

Solution architecture includes thorough  and robust systems for;

  • Content ingestion : High Throughput parsers/spiders for content acquisition, transformation and standardization
  • Semantic enrichment : Machine Learning models and taggers to semantically enrich the metadata with relevant concepts, making the data discoverable, reusable and traceable
  • Query parser : A natural language query parser to identify concepts, contexts and intentions from user keyword(s) and translate it to a query language that can be used to retrieve content from data lake(s)
  • Mail scheduler and user management parallels : A secure, fast and traceable system that can be configured to set up and manage actual alerts to subscribers 
  • Analytics and Decision aids : Integrated usage reports and infographics to assess the performance of the system, both qualitatively and quantitatively

 

Together these five components, heavily backed by AI modules make for an excellent, up-to-date, scalable and quick to integrate and feedback inclusive solution for the 

What gives this solution an edge?

  • Plug and Play content acquisition systems
  • Semantic enrichment with no prior/custom requirement and a provision to plug in custom ontologies/CVs if necessary
  • Scalability – Around 3500 users per week* 
  • Intelligent search and retrieval with due diligence to personalization – With little to no additional inputs from users
  • Integrated workflow components allowing decision support and feedback ingestion
  • Generic core modules that can be easily integrated with existing systems and/or adapted for a wide range of other applications

More Services

Semantic Enrichment

Knowledge Graphs

Ontology Engineering