AI: Sentiment Detection with BERT
BERT (Bidirectional Encoder Representations from Transformers) technology is revolutionizing the deep learning models (e.g., Google translate) and enabling transfer learning across all languages, such as transferring accurate sentiment, topic/ trend models from one language to another. eBrandValue takes this technology a step further using its Terabytes of available labeled data and know how that covers a decade in unstructured data processing and ever so dynamic social media. The incorporated infrastructure uses an ongoing learning and training with real-time user-defined data re-labeling and fast re-training features thanks to its extensive cloud-based resources.
AI: Label Classification with BERT
This deep-learning feature identifies a specific topic in the received contents as part of its real-time data flow pipeline. The topics can be configured if requested by the users/ clients to come from a specific set. The system provides a flexible user-defined environment with ongoing learning and training features. For example, already trained topics can be captured and classified for real time data monitoring and reporting with automated alarm mechanisms. Broad topics and related subtopics can be simultaneously labeled by this ever so intelligent and learning classifier robot.
AI: Language Detection
State of the art deep learning technologies (like eBrandValue can now detect a posts/ author's language with high accuracy globally over 100+ languages. This leads to better insights and author-level enrichment opportunities as part of the eBrandValue’s algorithm. For example, increasingly, certain authors use multiple languages. Such author-level behaviors signal certain author-level traits and lifestyles.
Better insights with Mindshare Effectiveness
eBrandValue develops and maintains a key metric called mindshare based on its decade-long experience and proprietary algorithm in its brand health and score card generation tools. Mindshare is a part of sales-leading indicator with its sales elasticity tracked and reported in Score Cards. This new feature enables brands to track in real-time how their mindshare is effective in gaining brand affinity. This can lead to faster responses in pivoting the current campaigns in addition to stop, increase or decresse gas types of decisions.
|
Enhanced monitoring with Spike Alerts
The significant increases in data volume are now configurable in flexible user defined formats using interval, brands/ keywords, sentiment and percent change as variable fields. As usual in eBrandValue's platform, brand definition may include smart topics, superbrands or many other flexible categories.
Better insights with Mindshare Effectiveness
eBrandValue develops and maintains a key metric called mindshare based on its decade-long experience and proprietary algorithm in its brand health and score card generation tools. Mindshare is a part of sales-leading indicator with its sales elasticity tracked and reported in Score Cards. This new feature enables brands to track in real-time how their mindshare is effective in gaining brand affinity. This can lead to faster responses in pivoting the current campaigns in addition to stop, increase or decresse gas types of decisions.
Smart Location Search
eBrandValue’s smart location search feature enables Geosearch based on extensive geo location map names (i.e., Central Park, Abbasağa Parkı, Cipriani Manhattan, Kadıköy, Asmalı Mescit, Uptown, Eskişehir, Ege Bölgesi). The smart search feature uses Google API to query the relevant enclosing latitude/ longitude polygon data points and provides all contents falling in the relevant area. For example, someone producing content from Hyde Park, or Çeşme or checked in a bar in Greenwich village will be automatically included in geo searches with the relevant semantic area search text, i.e., London, Ege Bölgesi, Manhattan, NY, US etc.
Location Enrichments
An author's location may come from bio, from the author’s posts, or pictures. In general, the user-generated content and engagement behavior provide many location-relevant information. eBrandValue uses even the keywords i.e., Eiffel Tower, Bebek Park, to identify an author's location in the most general and comprehensive sense. The underlying smart algorithm enables imputing a location for an author with all the data available in the author generated contents.
Author Enrichment
The past experiences show that the Author personalities and lifestyles may be identified using online behavior. For example, author interests, activities, opinions, and personality (i.e., OCEAN dimensions) are now all possible with this development with a smart/ learning infrastructure. Many smart applications become possible with this fundamental and crucial feature such as detecting fake news, segmentation, brand positioning and value proposition engineering.
Twitter API v2
eBrandValue has been an official partner of Twitter/GNIP for the last decade. This feature includes visual displaying and indexing in the database many new data fields provided by this Twitter API upgrade. These include Tweet Annotations, classifying topics and entities, Conversation IDs, Conversation control metadata, Poll results and poll options, also see https://developer.twitter.com/
To learn more about eBrandValue Platform, and how such metrics influence brand value, contact us using the form below.
