Web Page Audio Generation via Server-Side Text Filtering

The patent introduces an intermediary system comprising a server process and text-to-speech converter that mediates between the web page content and the user. The server process retrieves web page content, processes it through filtering to remove non-audible elements, converts it to speech audio, and delivers it to the user's client terminal. This intermediary architecture enables audio access to web pages without requiring complex client-side implementations, thus improving user convenience while maintaining manageable system complexity.

The patent replaces the mechanical interaction of reading text with the acoustic interaction of listening to speech. By substituting the visual-mechanical process of reading with an audio-based process, users gain the ability to access web page information through hearing, which improves accessibility and convenience particularly for users with visual impairments or those who prefer auditory learning.

The patent implements a multi-functional access system where a single interface provides multiple access modes for web page content. Users can select from text reading, audio playback, or a combination of both through a unified control mechanism. This universal access point eliminates the need to individually navigate and access each web page separately, as the system can process and deliver content in multiple formats through one interface, thereby improving time efficiency while maintaining operational simplicity.

The system performs preliminary processing of web page content by pre-converting text to speech audio and preparing multiple format versions before user request. This preliminary action allows users to immediately access content in their preferred format without waiting for conversion processes, significantly improving time efficiency when accessing pre-selected lists of web pages.

The patent inverts the traditional client-server architecture for text-to-speech conversion. Instead of placing the text-to-speech converter on the client terminal as in conventional approaches, the invention moves the conversion functionality to the server side. The server process retrieves web page content, converts it to speech audio, and delivers the audio file to the client terminal for playback. This inversion eliminates the need for complex client-side text-to-speech software, reducing client terminal requirements while maintaining access flexibility.

The patent extracts the text-to-speech conversion functionality from the client terminal and relocates it to the server process. By separating the conversion function from the client device, the system reduces the computational burden and software requirements on client terminals. The client terminal only needs basic audio playback capability, while the server handles the complex text processing and speech synthesis, thus reducing device complexity while preserving adaptability.

The patent extracts and removes non-audible information elements from web pages before converting the remaining content to speech. The server process filters out elements such as images, videos, advertisements, and other non-textual content that cannot be meaningfully converted to audio. This extraction process reduces the amount of content requiring text-to-speech conversion, thereby decreasing audio file sizes and transmission times while preserving all information that can be effectively communicated through audio.

The patent applies different processing qualities to different portions of web page content based on their suitability for audio conversion. Textual information is fully converted to speech with high fidelity, while non-textual elements are either excluded or represented through descriptive audio cues. This local quality approach ensures that information completeness is maintained for audio-appropriate content while optimizing transmission efficiency by excluding or simplifying representation of non-audio content.