TW201322018A - Method of analyzing personalized input automatically - Google Patents

Abstract

A computer-implemented method for analyzing personalized input automatically. The method uses one or more processors to capture an input content in the background. The input content is analyzed by said one or more processors to generate an output content, which is a personalized key word, phrase, or a combination of them. The output content is sent to a receiving terminal to retrieve personalized information.

Description

Automatic analysis of personalized input methods

The present invention relates to a method for automatically analyzing personalized input, and more particularly to a method for automatically analyzing personalized input in the background to generate output content, thereby obtaining a personalized message at the receiving end.

In the 21st century, the century of information technology, with the growth of electronic devices and information systems and the diversification of online information, the virtual and real society are closely integrated through these three. After the Internet has become one of the main ways of information dissemination, Internet users can receive a variety of messages in the 24-hour Internet world, regardless of location and time.

As the growth rate of information increases in equal proportions, information on news, communities, blogs, microblogs, etc. on the Internet is growing at an extremely fast rate. Since this information is quickly and abundantly flooded on the Internet, filtering and obtaining the required information from these astronomical information becomes a burden on the user, because filtering takes more time than obtaining a large amount of information.

Internet service providers provide services for searching or setting keyword terms for users to subscribe to information, and to find useful and useful information from astronomical information. However, such services have their actual implementation. The inconvenience and defects exist as follows:

1. Users are required to provide keywords or words according to their needs. If users do not provide keywords, they will not be able to obtain information that meets their needs.

2. These keywords or words are only relevant to the user's needs or meanings during a time period. After this period or after meeting the requirements, the information provided by these keywords or words becomes meaningless. Users need to update their keywords or words on their own, or they will continue to receive messages that have become non-compliant, causing problems.

The reason is that in order to solve the above-mentioned inconvenience, the relevant fields have not tried their best to find a solution, but the method that has not been applied for a long time has been developed. Therefore, how to obtain information that meets individual needs on the Internet more effectively is one of the current important research and development topics, and it has become an urgent target for improvement in related fields.

Accordingly, in view of the above various problems, the main object of the present invention is to provide a method for performing capture of input content in the background, and interpreting and analyzing the input content to generate an output content through which the output content is received from a receiving end. Get information that matches the output.

Another object of the present invention is to provide a method for interpreting content input by a person on a weekday basis, and finding a word or word commonly used by a person as a keyword or word to obtain service information that meets a specific needs of the individual.

Another object of the present invention is to provide an automatic and dynamic acquisition of personalized information from a receiving end without providing a keyword.

In order to achieve the above object, the present invention provides a preferred embodiment for automatically analyzing a personalized input for obtaining a personalized message at a receiving end, the method of using the one or more processors to perform the following steps: Automatically capturing an input content in a background; interpreting the input content by the processor; generating an output content; and transmitting the output content to the receiving end.

The step of interpreting the input content includes: loading an original word library; comparing the input content with the original word library to determine whether the corresponding original word can be found; and if so, inputting the original word frequency Comparing with a first set value to determine whether the input frequency is higher than the first set value; if not, ending the interpreting step. The step of generating an output content includes extracting an original word whose input frequency is higher than the first set value from the original word dictionary, and extracting the original word system includes a single word or a word or a plurality of words or words. set.

The step of interpreting the input content further includes: loading a learning word library; comparing the input content not included in the original word library with the learning word database to determine whether the corresponding learning word can be found. If yes, comparing the input frequency of the learning word with a second set value to determine whether the input frequency is higher than the second set value; if not, the input content is not included in the original word The word or word of the library is stored in the learning word library, and the input frequency of the stored learning word is compared with the second set value to determine whether the input frequency is higher than the second set value.

The step of generating an output content further includes extracting a learning word whose input frequency is higher than the second set value from the learning word dictionary, and the extracted learning word includes a single word or a word or a plurality of words or a collection of words. The output is a single keyword or a collection of words or a plurality of keywords or words.

The step of interpreting the input content further includes: comparing the input number of the learning words in the learning word database with a third set value to determine whether the input number is higher than the third set value; and inputting The learning words whose number is higher than the third set value are stored in the original word dictionary, and the input content includes a single word or a word or a plurality of words or a collection of words.

The above method further comprises a filtering step of analyzing the input content to filter the 赘 word or the word, and the 赘 word or word system includes a sensory verb, a degree adverb, a mediation word, a conjunction word, a pronoun or a function word. Or additionally including an association analysis step, the association analysis is performed on the output content. The automatic capture system further specifies one or more time ranges and captures the input content within the time range. The method further includes: obtaining a feedback information from the receiving end, and further presenting the feedback information.

In order to further understand the features and technical contents of the present invention. The preferred embodiments are described below in order to explain the details of the present invention, and are illustrated by the accompanying drawings. The symbols mentioned in the description refer to the schema symbols.

The invention is a method for automatically analyzing personalized input, which is applied to obtain a personalized message of a receiving end.

Please refer to Figures 1 and 2 for a first preferred embodiment of the present invention. FIG. 1 is a flow chart of a method for automatically analyzing personalized input according to the present invention, which uses one or more processors to perform the following steps:

Step 110: automatically extract an input content in a background by using the processor;

Step 120: The input content is interpreted by the processor;

Step 130: Generate an output content.

Step 140: transmitting the output content to the receiving end;

Step 150: Obtain a feedback information from the receiving end

Step 160: Present the feedback information.

The input content in the foregoing step 110 is a word or a word or a plurality of words or a collection of words, and the input content can be achieved by using various input methods in a specific implementation manner, for example, inputting characters by using a hard keyboard or a virtual keyboard. The touch screen is used to input text by hand or voice input through a microphone, but the invention is not limited to the ones listed herein.

In this step 110, the automatic capture in the background uses the method to record the user's input content without the user's knowledge. In a specific implementation, a regular input action is performed by a hard keyboard, for example, when inputting content on a paperwork system, the processor can process the hard keyboard input sequence through the paperwork system while simultaneously Actively record the hard keyboard input sequence.

The foregoing implementation includes the following three methods: the first system starts a hypervisor, and uses the interrupt and trap mechanism to obtain a keyboard input sequence; the second system implements a new device at the core level of the operating system. The driver replaces the original device driver function to achieve the purpose of capturing the keyboard input sequence; the third system uses the program interface provided by the operating system (OS) or the plug-in input method to obtain the keyboard input sequence.

In addition, using the touch screen to input text by hand, virtual keyboard input characters or voice input through a microphone may also be achieved by at least one of the above three embodiments. The method described above can be applied to various operating systems (OS) such as the Microsoft Windows series, the Linux series, the Unix series, the Android series, or the Mac OS X series, but the present invention is not limited to the ones listed herein.

Further, step 110 can be performed within one or more specific time periods. For example, if the user only wants to personalize the input of the off-duty time as the object of automatic analysis, the specific time period can be set to 7:00 to 9:00 in the morning. The specific implementation manner determines whether the call automatically captures the code of the input content by executing a code, the user specifies one or more specific time periods, and compares the system time with the specific time period.

Please refer to FIG. 2 for a more detailed description of the above steps 120 and 130. The specific execution of the foregoing step 120 as shown in FIG. 2 includes the following steps:

Step 210: Load a raw word library. The original word library stores one or more original words, which are words or words or a plurality of words or words that are generally known or commonly defined in a word dictionary;

Step 220: Comparing the input content with the original word library to determine whether the corresponding original word can be found, and if so, executing step 230, if otherwise, performing step 240;

Step 230: The input frequency of the original word is compared with a first set value to determine whether the input frequency is higher than the first set value, if otherwise step 240 is performed, if yes, step 250 is performed;

Step 240: End the interpretation step;

Step 250: Extract the original word whose input frequency is higher than the first set value from the original word library to generate the output content (step 130), and the output content analyzes the input content (step 120) Produce a single keyword or a collection of words or plural keywords or words.

Referring to Figures 1 and 2, the foregoing step 140 transmits the output to the receiving end. The receiving end can be a database, a news subscription service server or a multimedia subscription service server, but the invention is not limited to the ones listed herein.

In this step, the output content is transmitted to the service interface provided by the receiving end through a telecommunication link such as a serial interface, a Bluetooth interface, an internet network or a wireless mobile network, such as a database keyword query. The keyword input interface of the interface, the news subscription service, or the keyword input interface of the multimedia subscription service, but the invention is not limited to the ones listed herein.

In addition, at least one of the three steps of step 120, step 130 and step 140 may be performed in the background, that is, the user is not actively informed when the step is performed.

The foregoing step 150 obtains a feedback information from the receiving end. In this step, the feedback information is provided by the receiving end, and the specific implementation is that the receiving end searches for the message conforming to the output content according to the output content (ie, the keyword or the word or a combination thereof) to the foregoing service interface. The message that is met is a personalized message, including text messages, web links or multimedia materials, but is not limited to the ones listed here. The personalized message is obtained through, for example, an e-mail, a short message, a multimedia message, a simple information aggregation file, a web page, or a combination thereof, but the invention is not limited to the ones listed herein.

The aforementioned step 160 presents the feedback information. The specific implementation in this step is an electronic device having a display function such as a computer, a handheld device, a media playback device, etc. (but the invention is not limited to the ones listed herein) to visualize the personalized message transmitted from the receiving end.

Referring to Figures 3A and 3B, which are the second preferred embodiment of the present invention, as shown in Figure 3A, the steps of the preferred embodiment are mostly the same as the first preferred embodiment described above. Therefore, the same steps and symbols are not described again, but the preferred embodiment differs from the foregoing preferred embodiment in that the steps of the preferred embodiment include:

Step 220: Comparing the input content with the original word library to determine whether the corresponding original word can be found, if yes, executing step 230, if otherwise, performing step 310;

Step 310: Load a learning term library. The learning word library stores one or more learning words, the learning words are basically not present in the original word library, usually a person name, a company name, a product model or a self-made word, but are not limited thereto. Listed here;

Step 320: Comparing the input content not included in the original word dictionary with the learning word library to determine whether the corresponding learning word can be found, if otherwise, performing step 330, if yes, executing step 340;

Step 330: Store the words or words in the input content that are not included in the original word dictionary in the learning word database, and become a new learning word;

Step 340: The input frequency of the learning word is compared with a second set value to determine whether the input frequency is higher than the second set value. Otherwise, the step 240 is performed to end the determining step, and if yes, step 350 is performed. ;

Step 350: Extract a learning word whose input frequency is higher than the second set value from the learning word library to generate the output content (step 130).

As shown in FIG. 3B, in the preferred embodiment, the learning word can be converted into a conventional or customary original word according to the number of times the learning word is taken, and the specific implementation manner is After the foregoing step 310, the following steps are performed:

Step 360: Align the input number of the learning words in the learning word database with a third set value to determine whether the input number is higher than the third set value;

Step 370: Deposit the learning words whose input times are higher than the third set value into the original word dictionary.

Please refer to FIG. 4A, which is a third preferred embodiment of the present invention. As shown in the figure, most of the steps of the preferred embodiment are the same as the first and second preferred embodiments described above, and The same steps and symbols are repeated, but the preferred embodiment differs from the first and second preferred embodiments in that the steps of the preferred embodiment further include:

Step 410a: analyzing the input content to filter the 赘 word or word; and then, through the foregoing step 120, interpreting the filtered input content.

The specific implementation of the step 410a can establish a database of word words, and then analyze the input content and the vocabulary database to filter out the 赘 words or words. The above-mentioned words include sensory verbs, degree adverbs, mediators, conjunctions, pronouns or function words, etc., but the invention is not limited to the ones listed herein.

Please refer to FIG. 4B, which is a schematic diagram of another embodiment of the third preferred embodiment of the present invention. As shown in FIG. 4B, the difference from FIG. 4A is: another step 410b is performed to analyze the output content. Filtering the word or word; and transmitting the filtered output to the receiving end via the foregoing step 140.

The specific implementation of the step 410b is the same as the foregoing step 410a, and details are not described herein again.

Referring to FIG. 5, which is a fourth preferred embodiment of the present invention, as shown in the figure, most of the steps of the preferred embodiment are the same as the first and second preferred embodiments described above, and The same steps and symbols are described, but the preferred embodiment is different from the foregoing first and second preferred embodiments in that the steps of the preferred embodiment further include:

Step 510: Perform correlation analysis on the output content.

The specific implementation of the step 510, for example, one of the output keyword words is a vehicle brand, and the other keyword word is “fuel consumption”. According to the correlation analysis, the two keyword words have a considerable correlation. Sex, so it can be treated as the same set of output and sent to the receiver to get personalized messages. The association analysis implementation method can use the Association Rule Learning of data exploration to pre-establish the correlation of the word content as the basis for the association analysis after extracting the keyword words.

Referring to FIG. 6, which is a fifth preferred embodiment of the present invention, as shown in the figure, most of the steps of the preferred embodiment are the same as the first preferred embodiment described above, and the same is no longer the same. The steps and symbolic descriptions of the preferred embodiment are different from the foregoing first preferred embodiment. The steps of the preferred embodiment include:

Step 610: Give a score of the word or word according to the input frequency of each word or word of the output content;

Step 620: Associate each word or word of the output content with at least one of a plurality of categories;

Step 630: Transfer the output content, its corresponding score, and its associated category to the receiving end.

In step 610, the higher the input frequency is, the higher the score is. In an embodiment, the input frequency is further calculated according to different input time periods, and corresponding weights are assigned to each input time period, thereby calculating a weighted score of each word or word of the output content, and transmitting the weighted score to the Receiving end. Furthermore, the definition of the input time period, in addition to being divided according to absolute time, can also be defined for the individual relative time zone, for example, 9:00 to 6:00 am is defined as the working time, or 7:00 to 11:00. Points are defined as leisure time, but the way they are defined is not limited to the ones listed here. Step 620, by comparing with a related word classification database, to determine the category (such as science, entertainment, food, etc.).

In this embodiment, after the receiving end obtains the relevant content of the output content, it can be statistically and classified, and then generate a corresponding user database, which can be applied to the foregoing services, and can also be applied. For community services. For example, a social networking website can extract the input content of each user's high scores through the above-mentioned database, and perform analysis and comparison, thereby determining the correlation between the users, so as to provide feedback information of the friend's suggestion. The determination of the inter-user relevance may be further limited to one or more specific categories.

In addition, in another embodiment, the output content may be generated by other different user input methods and analysis methods, and is not limited to the foregoing embodiment.

With the above implementation of the present invention, the user does not need to input or update the keyword words for the relevant information service, but uses an automatic analysis method of personalization to collect personalized keyword words to dynamically obtain the user. Meaningful personalized messages.

While the present invention has been described above in terms of the preferred embodiments thereof, it is not intended to limit the invention, and the invention may be modified and modified without departing from the spirit and scope of the invention. The patent protection scope of the invention is subject to the definition of the scope of the patent application attached to the specification.

110~160. . . step

210~250. . . step

310~370. . . step

410a and 410b. . . step

510. . . step

610~630. . . step

Figure 1 is a flow chart showing the main steps of the first preferred embodiment of the present invention;

Figure 2 is a flow chart showing the detailed steps of the first preferred embodiment of the present invention;

3A and 3B are flowcharts showing steps of adding a learning word dictionary to the second preferred embodiment of the present invention;

Figure 4A is a flow chart showing the steps of filtering the word filtering according to the third preferred embodiment of the present invention;

Figure 4B is a flow chart showing the steps of adding another 赘 word filtering step according to the third preferred embodiment of the present invention;

Figure 5 is a flow chart showing the steps of adding correlation analysis according to the fourth preferred embodiment of the present invention;

Figure 6 is a flow chart showing the steps of adding the output content attribute recording step in the fifth preferred embodiment of the present invention.

110~160. . . step

Claims (22)

A method for automatically analyzing personalized input is applied to obtain a personalized message of a receiving end, wherein the method uses one or more processors to perform the following steps: automatically extracting an input content in a background by using the processor; Determining the input content by the processor; generating an output content; and transmitting the output content to the receiving end. A method of automatically analyzing personalized input as described in claim 1 wherein the input includes a single word or a plurality of words or a collection of words. The method for automatically analyzing personalized input as described in claim 2, further comprising: giving the word or word a score according to an input frequency of each word or word of the output content, and transmitting the score to the receiving end. A method for automatically analyzing personalized input as described in claim 3, wherein the word or word score is given according to an input frequency of each word or word of the output content, and the score is transmitted to the receiving end. In step, the input frequency is separately calculated according to different input periods, and corresponding weights are assigned to each input period, and then the weighted score of the output content is calculated, and the weighted score is transmitted to the receiving end. The method for automatically analyzing personalized input as described in claim 2, further comprising: associating each word or word in the output content with at least one of the plurality of categories, and transmitting the category of the association To the receiving end. The method for automatically analyzing personalized input as described in claim 1, wherein the step of interpreting the input includes: loading an original word library; comparing the input content with the original word library to determine whether Find the corresponding original word; if so, extract the corresponding original word from the original word library. The method for automatically analyzing personalized input as described in claim 6, wherein the step of comparing the input content with the original word library to determine whether the corresponding original word can be found includes the corresponding The input frequency of the original word is compared with a first set value to determine whether the input frequency is higher than the first set value. A method for automatically analyzing personalized input as described in claim 6 wherein the extracted original word comprises a single word or a plurality of words or a collection of words. The method for automatically analyzing personalized input as described in claim 6 or 7, wherein the step of interpreting the input further comprises: loading a learning word library; the input content is not included in the original word. The word or word of the library is stored in the learning word library; the input frequency of the learning word is compared with a second set value to determine whether the input frequency is higher than the second set value; if so, the corresponding The learning words are extracted from the learning word library. A method of automatically analyzing personalized input as described in claim 9 wherein the extracted learning term comprises a single word or a plurality of words or a collection of words. The method for automatically analyzing personalized input according to claim 9 of the patent application, wherein the step of reading the input content further comprises: comparing the input number of the learning words in the learning word database with a third set value And determining whether the input number is higher than the third set value; and storing the learning words whose input times are higher than the third set value into the original word dictionary. The method for automatically analyzing personalized input as described in claim 1 of the patent application further includes a filtering step of analyzing the input content to filter the 赘 word or word. The method for automatically analyzing personalized input as described in claim 1 of the patent scope further includes a filtering step of analyzing the output content to filter the word or word. The method for automatically analyzing personalized input as described in claim 12 or 13, wherein the 赘 word or word system includes a sensory verb, a degree adverb, a preposition, a conjunction, a pronoun or a vocabulary. The method for automatically analyzing personalized input as described in claim 1 of the patent application further includes an association analysis step for performing correlation analysis on the output content. The method of automatically analyzing personalized input as described in claim 1, wherein the automatic capture further specifies one or more time ranges and captures the input content within the time range. The method for automatically analyzing personalized input as described in claim 1 further includes: obtaining a feedback information from the receiving end; and presenting the feedback information. The method for automatically analyzing personalized input as described in claim 1, wherein the input content is interpreted; an output content is generated, and at least one of the three steps of transmitting the output content to the receiving end is tied to a background. Take action. The method of automatically analyzing personalized input as described in claim 1, wherein the output content is a single keyword or a plurality of keywords or a collection of words. A method for automatically analyzing personalized input is used to analyze the relevance of input content between more than one user. The method uses one or more processors to perform the following steps: providing input content of more than one user; Each user gives a corresponding score to the word or word according to the input frequency of each word or word in the input content; for each user input content, extracts a word or word whose score is higher than a preset value And analyzing the relevance of input between different users based on the extracted words or words. The method for automatically analyzing personalized input as described in claim 20, wherein each user is given a corresponding score for the word or word according to the input frequency of each word or word in the input content. The input frequency is separately calculated according to different input periods, and corresponding weights are assigned to each input period, and then the weighted score of the word or word is calculated, and extracted according to the weighted score. The method for automatically analyzing personalized input as described in claim 20, further comprising: for each user, each word or word in the input content is associated with at least one of the plurality of categories. And the extracting step is performed for at least one of the plurality of categories.