TWI766457B - Analysis system and upload method for language - Google Patents

Abstract

An analysis system and method for language includes a login module for users to log in. The login module is electrically connected to a collection module that is used to collect and manage various research corpora and conduct management operations through webpage. The management operations include modification, analysis and deletion procedures. The collection module is electrically connected to a transcription module, which is used to the uploaded corpus, audio and video data is used for transcoding, and the relevant information fields of the file to be entered are also displayed on the same page. In addition, the transcription module is electrically connected to an analysis module. It is used to analyze the transcribed corpus files. The analysis module has the function of word hyphenation and analyzes language samples in an automated manner.

Description

Analysis system and upload analysis method for language

The present invention relates to an analysis system, in particular to a language analysis system integrating uploading, transcription and analysis functions.

Children's language development is an important indicator in the development of cognitive ability, which will affect children's learning and thinking performance. According to domestic research, it is found that the incidence of language disorders is higher at a young age, and language disorders have the highest incidence among special education objects. If children's language development problems can be detected and treated at an early stage, the incidence of language barriers can be reduced to avoid social problems and emotional distress caused by language barriers.

Children's language development can be recorded by audio or video recording of children's daily language use, and then the speech uttered by the child is transcribed into text files by a speech therapist. Finally, language samples are analyzed through a computer system to calculate the indicators related to children's language development. Have a speech therapist do an assessment to find out where the child's language development is.

Over the past few years, the existing language analysis system for Chinese-speaking children has provided the following functions, including formulating a set of standard language transcription formats; providing analysis programs for language development index analysis; providing web page sharing services and other functions; however, the existing systems have the following problems: Including that the existing system has not been fully integrated, there is no consistent interface in use, and there are complex and rigorous regulations in the developed transcription format. Therefore, it takes time to learn when it is used for the first time. In addition, there are still many language analysis systems in different formats on the market, and it takes time and cost to convert. There are problems such as extra word segmentation actions when the system is used.

Therefore, there are still many inconveniences in the analysis of Chinese language samples for child language therapists in the existing system. How to improve and simplify the intermediate steps is an important issue.

In view of the above deficiencies, the main purpose of the present invention is to provide a language analysis system and upload analysis method, import SaaS (Service as a service) service into the analysis of children's language samples, and use web technology to analyze three steps. The functions required for collection, transcription, and analysis are integrated into one, and a single-window interface can be implemented to complete the child language sample analysis system, saving researchers the time and complexity required for child language sample analysis.

In order to achieve the above-mentioned purpose, the present invention mainly provides a language analysis system, which includes a login module, which is used for user login; and a collection module, which is electrically connected to the login Module, the collection module is used to collect and manage the corpus of various studies. The collection module further includes a corpus unit, which is used for corpus management. The corpus unit displays various uploaded data through the rows. The corpus data is managed through web pages, and the management operations include modification, analysis and deletion procedures; a corpus recording unit is electrically connected to the corpus unit to store the uploaded corpus data; a language A corpus adding unit is electrically connected to the corpus unit, and the corpus adding unit is used to start a program for uploading new corpus; a corpus sharing unit is electrically connected to the corpus unit, and the corpus is shared The unit is used for a single user to share the uploaded corpus information; a transcription module is electrically connected to the collection module, and the transcription module is used for transcoding the uploaded corpus audio and video data. , the transcription module further includes an uploading unit, which is automatically connected after the new unit of the corpus is activated The uploading unit uploads the video and audio data to be uploaded by dragging or opening the file, and simultaneously displays the relevant information fields of the received file on the same page; a transcription unit is electrically connected the uploading unit, the transcription unit is used for assisting the transcription of the uploaded semantic data; an analysis module is electrically connected to the transcription module, the analysis module is used for analyzing the transcribed corpus file, the The analysis module further includes: a corpus analysis unit, the corpus analysis unit is used for analyzing the voice file, the corpus analysis unit has the function of hyphenation and word segmentation, and analyzes the language samples in an automated manner; and a The exporting unit is electrically connected to the corpus analysis unit, and the exporting unit is used for receiving and outputting the analysis data from the corpus analyzing unit.

Another object of the present invention is to provide an upload analysis method for language, which includes selecting upload data (S1); transcribing according to the upload data type (S2); if the upload data type is either audio or video, then A transcription interface is provided for transcription (S3); if the uploaded file is in a specified file format, the file format is parsed and transcribed (S4); after uploading and transcription is completed, word segmentation and analysis are performed on the transcribed file (S5); output The analysis result (S6).

In order to make the above-mentioned and other objects, features and advantages of the present invention more clearly understood, preferred embodiments are exemplified below, and are described in detail as follows in conjunction with the accompanying drawings.

1: Login to the module

11: Member registration unit

12: Member login unit

2: Collect modules

21: Corpus Unit

22: Corpus Recording Unit

23: New unit of corpus

24: Corpus Sharing Unit

3: Transcription module

31: Upload unit

32: Transcription unit

4: Analysis module

41: Corpus Analysis Unit

42: Export unit

The first figure is a system block diagram of the present invention.

The second figure is a block diagram of the collection module of the present invention.

The third figure is a block diagram of the transcription module of the present invention.

The fourth figure is a block diagram of the analysis module of the present invention.

The fifth figure is a flow chart of the upload analysis method of the present invention.

Please refer to the first figure, which is a system block diagram of the present invention. The language analysis system of the present invention mainly includes a login module 1, a collection module 2, a transcription module 3 and an analysis module 4, wherein the login module 1 is used for user login. , the login module 1 further includes a member registration unit 11 and a member login unit 12. The member registration unit 11 is provided for the first-time user to input information and register as a member, and use this as a member's The data is stored in the system, and the member login unit 12 is used to log in to the system to compare whether a member has been registered.

Continue to refer to the first figure. The login module 1 is electrically connected to the collection module 2, and the collection module block diagram of the second figure is referred to. The collection module 2 is used to collect and manage various research corpora. The collection module 2 further includes a corpus unit 21, the corpus unit 21 is used for corpus management, the corpus unit 21 displays various uploaded corpus data through rows and columns, and performs management operations through web pages. Including modification, analysis and deletion procedures; the corpus unit 21 is electrically connected to a corpus recording unit 22, and the corpus recording unit 22 is a database for storing uploaded corpus data; and, the corpus unit 21 is electrically connected to a corpus adding unit 23 and a corpus sharing unit 24. The corpus adding unit 23 is used to start a program for uploading a new corpus. In this embodiment, the corpus adding unit 23 The operation is performed through a web page, and the corpus sharing unit 24 is used for a single user to share the uploaded corpus information.

Continue to refer to the first figure. The collection module 2 is electrically connected to the transcription module 3, and the transcription module 3 is used for converting the uploaded corpus audio-visual data. The transcription module 3 further includes an uploading unit 31, as shown in the first As shown in the block diagram of the transcription module in Figure 3, after the corpus adding unit 23 is activated, the system automatically connects to the uploading unit 31, and through the web-based operation, the video and audio data to be uploaded can be dragged or opened by any method. One way to upload, and also display the received file to be input on the same page The relevant information fields of the case, in this embodiment, these fields include name, age, gender, recording location, situation introduction, participants and the date of receipt of the case and other information, and these information fields also have different language samples. Analyze the required information for design, including adding web page interactive functions such as speaker, sentence start and end time, sentence content, and sentence annotation; in addition, the uploading unit 31 is electrically connected to a transcription unit 32, and the transcription unit 32 uses For the purpose of assisting the transcription of the uploaded semantic data, in this embodiment, the transcription unit 32 performs transcription assistance through a speech recognition model, such as the Google speech recognition model. In the uttered section of the video, the uttered section is cut, and then the speech recognition model is used for speech recognition; in addition, the transcription module 3 is that the transcribed corpus file of the present invention can be in line with traditional language sample analysis tools. Module 3 has the function of supporting the export of transcription completed sentences into a standard or format conforming to the tool.

See the first image. The transcription module 3 is electrically connected to the analysis module 4, and the analysis module 4 is used to analyze the transcribed corpus file. The analysis module 4 further includes a corpus analysis unit 41, as shown in the fourth figure As shown in the block diagram of the analysis module, the corpus analysis unit 41 is used to analyze the voice file and the transcribed file. The corpus analysis unit 41 has the function of hyphenation and hyphenation, and performs language sample analysis in an automated manner. , the analysis content includes calculating the language development indicators of the sentences in other files of the file content. These indicators include analysis indicators such as part-of-speech ratio, average sentence length, etc. The results of the analysis indicators can be visualized on the webpage. , to increase the readability of the indicators, and combined with text mining technology to display information in different aspects; the corpus analysis unit 41 is electrically connected to an export unit 42, and the export unit 42 is used for receiving data from the corpus analysis unit. 41 analysis data, and use either CHILDES or CLSA format as the export format.

Please refer to Fig. 5, which is a flowchart of the upload analysis method of the present invention. The uploading analysis method of the present invention includes selecting uploading data (S1), and the type of data that can be uploaded includes any one of a specified file format, audio or video; transcribing according to the uploading data type (S2), wherein the uploading data type is audio or any of the images, a transcription interface will be provided for transcription (S3), the transcription interface It is the input of field data for the audio and video files, the field data includes the speaker, the start and end time of the sentence, the content of the sentence and the comment of the sentence, etc., as well as the personal data input of the audio and video files, including name, age, gender, recording location, If the uploaded file is in the specified file format, the file format is parsed and transcribed (S4), including extracting the roles, sentences, time and other information listed in the file, and Transcribe; after uploading and transcribing, perform word segmentation and analysis on the transcribed file (S5), the analysis content includes calculating the language development indicators of the sentences in other files of the file content, and these indicators include the proportion of parts of speech , average sentence length and other analysis indicators; finally, output the analysis result (S6), the analysis result can use any one of the two specified CHILDES and CLSA formats as the export format, and at the same time use the visual way to present on the web page.

However, the above-mentioned embodiments are preferred implementation examples, which should not limit the scope of the present invention. Any equivalent changes or modifications made according to the scope of the patent application of the present invention and the contents of the description shall belong to the following aspects of the present invention. the scope of patent coverage.

1: Login to the module

11: Member registration unit

12: Member login unit

2: Collect modules

3: Transcription module

4: Analysis module

Claims (8)

An analysis system for language, comprising: a login module, which is used for user login; a collection module, which is electrically connected to the login module, and the collection module is used for Collection and management of various research corpora, the collection module further includes: a corpus unit, which is used for corpus management. Perform management operations, which include modifying, analyzing and deleting programs; a corpus recording unit, which is electrically connected to the corpus unit for storing uploaded corpus data; a corpus addition unit, which is electrically connected The corpus unit, the corpus adding unit is used to start the program for uploading new corpus; a corpus sharing unit is electrically connected to the corpus unit, and the corpus sharing unit is used for a single user to upload the uploaded corpus. The corpus information is shared; a transcription module is electrically connected to the collection module, and the transcription module is used for converting the uploaded corpus audio-visual data. The transcription module further includes: a The uploading unit is automatically connected to the uploading unit after the corpus addition unit is activated, and uploads the audio and video data to be uploaded to the analysis system by either dragging or opening the file, and simultaneously displaying the desired input on the same page The relevant information field of the received file; a transcription unit, which is electrically connected to the uploading unit, and the transcription unit is used to assist in the transcription of the uploaded semantic data; an analysis module, which is electrically connected to the transcription module group, the analysis module is used to analyze the transcribed corpus file, and the analysis module further includes: a corpus analysis unit, the corpus analysis unit is used for analyzing the voice file and the transcribed file, and the corpus analysis unit has the function of hyphenation and word segmentation, so as to analyze the language samples; and an export unit, The corpus analysis unit is electrically connected, and the export unit is used for receiving the analysis data from the corpus analysis unit and outputting it to an electronic device. The analysis system as claimed in claim 1, wherein the fields include name, age, gender, recording location, situation introduction, participants, and date of receipt, etc., and the information fields also have analysis for different language samples The required information is designed, including the addition of the speaker, the start and end time of the sentence, the content of the sentence and the annotation of the sentence. The analysis system according to claim 1, wherein the export format is any one of CHILDES and CLSA formats as the export format. A method for uploading and analyzing language, comprising: a. selecting uploading data through an uploading unit (S1); b. transcribing according to the uploading data type through a transcription unit (S2); c. the uploading data type If it is either audio or video, a transcription interface will be provided for transcription (S3); d. If the file of the uploaded data is in the specified file format, the file format will be parsed and transcribed (S4); e. After uploading and transcribing, Perform word segmentation and analysis on the transcribed file through a corpus analysis unit (S5); f. output the analysis result through an export unit (S6) to an electronic device. The upload analysis method according to claim 4, wherein the type of the uploaded data in step a includes any one of a specified file format, audio or video. The upload analysis method according to claim 4, wherein in step c, the transcription interface is to input field data for the audio and video files, and the field data includes the speaker, the start and end time of the sentence, the content of the sentence, and the annotation of the sentence, etc., And the input of personal data of the audio and video files, including name, age, gender, recording location, situation introduction, participants and the date of receipt of the case. The uploading analysis method according to claim 4, wherein in step d, information such as roles, sentences, and time listed in the file is extracted and transcribed. The upload analysis method according to claim 4, wherein in step f, the analysis result can use any one of the two specified CHILDES and CLSA formats as the export format.

Images