Skip to content

Overview

Info

This document is a work in progress.

Telenav, a leading provider of connected-car and location-based services, provides semantic search for on-board (Edge), off-board (Cloud), and hybrid (Edge/Cloud). Telenav delivers these leading-edge search services to support automotive navigation services for major automotive partners.

Powerful semantic and location aware search delivers highly relevant and useful results to the driver. Results are based on understanding user’s intent and context using multiple criteria including location, day/time, previous search history, etc.

Entity Service, enables developers to build powerful search features for a variety of applications. In addition to powerful onebox search, the features include query prediction (auto suggestion / word prediction), entity detail lookup, and discovery features. The Telenav cloud services exposed by the Entity Service REST API, enable developers to build powerful and engaging search based applications.

Search is the backbone of the service and the search methods enable the developer to build all of the below search (and related) features taking advantage of the Telenav search technology and Telenav cloud. "Location Based Telenav Semantic Search", using a powerful contextual and flexible search engine, finds POIs (points of interest) and addresses using free text or tagged query syntax. Advanced semantic knowledge and underlying search technology provide highly relevant search results by simply entering a free text query in a single text field (one-box).

Advanced query parsing algorithms, with help of robust global semantics DB enables recognition of user’s intent by interpreting the multiple search query patterns across multiple geographies and languages. This tagged query syntax also enables developers to build voice search use cases by mapping voice to text NLU output, to a structured query input.

Important capabilities of Location Based Telenav Semantic Search include:

  • Semantic based free text query search recognition
  • Pattern combinations include 'What', 'Where', and 'What in Where' combinations
  • Intelligent parsing recognizes addresses, streets, POIs, categories, intersections, and cities as input types.
  • Discovery features of category search and nearby categories details
  • Rich content (i.e. ratings) available for some POIs
  • Specialized search features for categories such as electric vehicle charge stations, parking, and fuel stations
  • Multi box search supported
  • Data entry reduction by automatic suggestions based on few characters of input
  • Word prediction using patented technology to propose likely word input
  • System learns users behavior and interactions to personalize and continuously improve search experience

With Telenav cloud services exposed by the Entity Service REST API, a developer can build search features to access millions of entities such as POIs, addresses, streets, categories, cities, etc.

Refer to Search section for details.

Predictions

The predictions method enables developers to build auto suggestion type capabilities which greater reduce user text data entry. These methods provide access to Telenav's auto-suggest methods and proprietary word prediction capabilities.

  • Suggestion Prediction

Assists user to find intended result by ‘suggesting’ likely search targets with limited keystrokes. Results provided are based on location, popularity and user's History. Autosuggest recognizes many types of entities including addresses, POIs, brands, streets, and cities. POIs of similar brand are grouped into a single suggestion. Results are optimized by location, prominence, popularity, and history.

  • Word Prediction

A patented solution to dramatically speed up search entry by suggesting words the user is intending to enter based on factors such as what’s nearby, past driver behavior, and driver content.

Refer to Prediction section for details.

Discover

Thorough taxonomy of useful categories and sub-categories, Discover feature enables developer to fetch category tree, explore nearby categories, fetch places of a specific category.

Refer to Discover section for details