TORONTO, April 10, 2012 /CNW/ - Sprylogics International Corp. - TSX Venture: SPY ("Sprylogics" or the "Company") is pleased to provide a technology update:
The Sprylogics Technology Team is focused on leveraging and enhancing the existing patent pending technology base to build mobile search solutions that interpret what people are saying online and on their mobile phones in order to:
a) Better understand what they are looking for (query intent);
b) Better understand trends and patterns in people's behavior and opinions in aggregate (improve quality and relevancy of search results).
This is accomplished through the use of semantic technologies and natural language processing techniques like entity extraction, semantic graph creation, disambiguation, matching and clustering to process massive volumes of unstructured data in order to extract key sentiments, facts, opinions, user interests and intents.
The team which is comprised of multiple engineers, data scientists, semantic researchers and mobile application developers and includes 3 PhDs is focused on four streams of development activities.
1. Named Entity, Fact, and Opinion Extraction:
The Sprylogics Natural Language Processing team has created an Software Development Kit ("SDK") that enables partners to extract entities, facts and opinions out of unstructured text. The SDK uses a combination of statistical techniques (machine learning trained algorithms) and formal grammar to analyze sentence structure, identify participating entities and extract relationships between sentence elements that signify facts and opinions.
The extraction team has recently expanded the core Natural Language Processing capabilities of the platform through improved precision and accuracy of entity, fact and opinion extraction and has recently distributed an updated version of the Natural Language extraction SDK to select partners.
2. Machine Learning:
The Sprylogics Machine Learning team has evaluated and implemented various machine learning algorithms for the purposes of text classification. Trained algorithms are able to classify sentences in natural format as they occur in common conversations, such as in chats, on Facebook or natural language search queries into multiple verticals which enable querying the most appropriate APIs for obtaining answers. The algorithms developed by the Machine Learning team are currently able to classify sentences that occur in natural conversations for multiple key verticals of interest with mid 90% accuracy.
3. Knowledge Base:
The Sprylogics Knowledge Base team has concentrated on creating a fast, distributed and scalable knowledge base infrastructure which stores known facts about millions of entities. The Knowledge base team has integrated multiple data sources including dbpedia and others into the core processing pipeline to further enable discovery of facts about extracted entities.
4. Mobile Development and APIs:
The advances in the improvement of the core natural language processing and semantic platform have culminated in the creation of the APIs and mobile development SDKs that enable partners to embed intuitive, semantically relevant search and advertising experience into their mobile applications. The Mobile Development team has successfully demonstrated its "Liquid Messaging" Search API and HTML5 interfaces to several leading messaging providers and is now entering technical evaluation and proof of concept phases with select partners.
The technology team is currently on track for the upcoming early June milestone of delivering the initial production release of the APIs and mobile SDKs to various partners for integration into mobile messaging applications.
"I am very excited with the immense progress that we have made in our objective of better understanding of natural language thereby improving our semantic search capabilities. The extraction, processing and interpretation of natural language at scale represents some of the most challenging technical problems today, but with our extremely talented research and technology team, we continue to advance our patent pending technologies at a rapid pace. We look forward to the integration of our solutions into our mobile partner applications which will enable a very differentiated mobile search user experience which can be easily monetized," said Alex Zivkovic, CTO of Sprylogics.
About Sprylogics International Corp.
Sprylogics International develops advanced search and analysis technology. The patent pending technology platform uses Semantic and Machine Learning techniques to process, analyze and interpret unstructured data including real time conversation, in order to extract key sentiments, facts, user interests and intent. The technology platform is currently being used to develop innovative solutions related to search and discovery of products and places on mobile devices. Find out more at www.sprylogics.com.
Neither TSX Venture Exchange nor its Regulation Services Provider (as that term is defined in the policies of the TSX Venture Exchange) accepts responsibility for the adequacy or accuracy of this release.
Cautionary and Forward-Looking Statements
Statements contained in this news release which are not historical facts are forward-looking statements that involve risk, uncertainties and other factors that could cause actual results to differ materially from those expressed or implied by such forward-looking statements. There can be no assurance that such statements will prove to be accurate and actual results and future events could differ materially from those anticipated in such statements. All forward-looking statements included in this news release are based on information available to the Company on the date hereof. The Company assumes no obligation to update any forward-looking statements, except as required by applicable securities laws.
For further information:
Sprylogics International Corp.
David Berman, CFO, 905-761-9200 Ext.21 or [email protected]
Share this article