Speech recognition software enables users to enter commands and perform actions on their computers using standard microphones and a predefined series of verbal commands. This allows users to open programs and input text strictly by using their voices. Open source speech recognition programs allow end users to view and potentially modify the source code for programs. Modification of the source code allows changes to be made to all features of the software, enabling skilled users to add, remove, and otherwise modify features on the software.
The primary advantage to open source design in open source speech recognition programs is that it provides end users with a greater level of customization than a traditional closed source program would. In closed source programs, the actual program code is both locked and hidden from the end user, preventing any changes from being made. A closed source speech recognition program provides options for changing the keywords used to access various features. Thus, an open source speech recognition program does away with the limitations of the programming language over commands; by reconfiguring the source code, the end user can theoretically alter the keywords for the program to anything they can imagine.
Another benefit of open source speech recognition is that it allows for collaborative customization of software. When other end users make useful modifications, they will often link their versions of the program online, allowing others to use the tools they have added or modified on the software. This means there can be an endless number of variations on the software. One could contrast this with a closed source speech recognition program, where the only possible version is the version issued by the manufacturer.
One of the largest disadvantages to open source speech recognition programs is that they cater to the technologically savvy. Without a background in programming, it can become overwhelming to modify even the simplest features on the program. Additionally, it can be hard to tell which versions of the program to “trust,” as both honest and unscrupulous users can modify the program equally. Some downloadable versions of the program might contain spyware, adware, or other malicious types of software. This is never a problem with closed source applications, as end users know they are receiving no more than what the software maker coded into the program.