Home  » Company » Technology

Technology

SocialNuggets technology developed over 20 man-years unearths critical customer sentiments and feedback embedded in millions of online conversations occurring over the social landscape. SocialNuggets technology converts free form textual content into structured contextual information that can be analyzed using any of the standard industry database access tools.

This technology has the following key components:

  • Focused Harvesting
  • Smart Connectors
  • SocialNuggets Engine
  • Data Output

Focused Harvesting

Using automated and curation techniques, we identify relevant sources for various vertical industries. The list of sources are constantly reviewed and modified to keep up with the changing landscape. Using various collection techniques, millions of engagements occurring in the social networks such as Facebook, Twitter, and industry-specific Forums, Reviews and Blogs are identified, extracted and stored so that advanced analysis and data modeling techniques can be applied.

Smart Connectors

We have developed over 200 connectors that enable us to extract the right information in the appropriate format from various online conversations. We have developed connectors for many sites using their APIs (application programming interfaces) such as for Facebook and Twitter. However, for many sites, especially review sites and forums, standard APIs may not be available and so we have developed proprietary connectors to understand and interpret data correctly from these sites.

SocialNuggets Engine

SocialNuggets Engine can be classified into following sub-headings

  • Natural Language Processing & Text Analytics
  • Domain and Industry-Specific Knowledge and Semantics
  • Customer Sentiment Analysis

Natural Language Processing & Text Analytics

Based on Natural Language Processing (NLP) techniques, the SocialNuggets technology uses a combination of statistical, linguistic and ontology-based approaches to establish a deeper understanding of the text. Our Text Analytics Engine automatically builds an accurate and comprehensive map of entities, facts and relationships in the data which then becomes the foundation for further processing with industry filters.

Domain and Industry-Specific Knowledge and Semantics

By focusing our efforts on specific industries, such as specific products like tablets and smartphones in Consumer Electronics industry, the output of the NLP/Text Analytics Engine is processed, filtered and mapped against the taxonomy of the specific product category. Not only does the domain establish focus for data extraction (which online sources to harvest), it significantly increases the ability of the technology to dig deep into the real meaning of the customer’s online posting and accurately establish intent.

While our current domain focus is Consumer Electronics, our technology has wider appeal for various industry segments.

Customer Sentiment Analysis

Armed with the output of the language processing matched against the taxonomy of the domain, SocialNuggets uses patent-pending, deep and high-accuracy sentiment analysis to determine customer sentiment. Built-in machine learning capability allows the sentiment analysis engine to be tweaked and tuned over time and increase accuracy and dependability. As domain-specific concepts and nuances are stored in the engine it provides highly accurate and actionable sentiments. For example, our engine can deduce that long battery life is positive while a long recharge time is negative.

Data Output & Reporting

We store meta data, sentiments and all relevant data about each of the posts analyzed. This giant database is stored in a Greenplum data warehouse format.

As part of subscription to our Premium Data Portal, this data can be customized/accessed using our Flex based portal, REST APIs, CSV files and any number of BI tools including Tableau, SAS or SPSS. We also take a subset of this data and provide nuggets in the form of infographics on our site.