CN104778202B - The analysis method and system of event evolutionary process based on keyword - Google Patents

Abstract

Embodiments of the present invention provide a method and system for analyzing event evolution process based on keywords. The method includes: counting the number of search results in each time period, and obtaining a sequence in which the number of search results changes with the time period; performing peak detection on the sequence to obtain at least one peak window; performing text analysis on the search results corresponding to the peak window to obtain Spike descriptions; show the sequence of the number of search results over time and show the spike descriptions at the locations of the spikes. In the embodiment of the present invention, by counting the sequence of the number of search results changing with the time period, displaying the sequence of the number of search results changing with the time period and the peak description, showing the trend of the number of microblog information changing with time, and the number of microblog information changing with time Trends can show the evolution process of events related to keywords. According to the peak description, users can know the main content of Weibo information without reading Weibo content in detail, which improves the amount of information displayed in search results.

Description

Analysis method and system of event evolution process based on keywords

technical field

The embodiments of the present invention relate to the field of computer technology, in particular to a method and system for analyzing event evolution process based on keywords.

Background technique

With the development of computer technology, microblogging software, as an important source of information acquisition, has become a widely used social software.

In the prior art, the microblog software performs a matching search through the keyword input by the user, and displays the searched microblog information closely related to the keyword on the user terminal.

Due to the time attribute of microblog information, the existing microblog information display results cannot show the trend of microblog information changing over time, resulting in a low amount of information displayed in the search results, and the evolution process of events related to keywords cannot be displayed .

Contents of the invention

Embodiments of the present invention provide a method and system for analyzing event evolution process based on keywords, so as to increase the amount of information displayed in search results and display the evolution process of events related to keywords.

An aspect of the embodiments of the present invention is to provide a method for analyzing event evolution process based on keywords, including:

Counting the number of search results in each time period, and obtaining the sequence of the number of search results changing with the time period;

Performing peak detection on the sequence to obtain at least one peak window, each of the peak windows includes a subsequence, the subsequence includes a spike, and the spike is a local maximum of the number of search results in the sequence ;

Perform text analysis on the search results corresponding to the peak window to obtain a peak description;

A sequence of changes in the number of search results with the time period is displayed, and a description of the peak is displayed at the position of the peak.

Another aspect of the embodiments of the present invention is to provide a keyword-based analysis system for event evolution process, including:

A statistics module, configured to count the number of search results in each time period, and obtain a sequence in which the number of search results changes with the time period;

A detection module, configured to perform peak detection on the sequence to obtain at least one peak window, each of the peak windows includes a subsequence, the subsequence includes a peak, and the peak is the number of search results in the sequence Local maximum in ;

A text analysis module, configured to perform text analysis on the search results corresponding to the peak window to obtain a peak description;

A display module, configured to display the sequence of the number of search results changing with the time period, and display the description of the peak at the position of the peak.

The method and system for analyzing the event evolution process based on keywords provided by the embodiments of the present invention obtains the local maximum value of the sequence by counting the number of search results that change with the time period according to the peak detection, and analyzes the cause of the local maximum value Obtain the peak description, display the sequence of the number of search results changing over time, and display the peak description at the peak position, and realize the microblog information display results can show the trend of the number of microblog information changing over time, and the number of microblog information changing over time Trends can show the evolution process of events related to keywords. According to the peak description, users can know the main content of Weibo information without reading Weibo content in detail, which improves the amount of information displayed in search results.

Description of drawings

Fig. 1 is the flow chart of the analysis method of the event evolution process based on keywords provided by the embodiment of the present invention;

Fig. 2 is a schematic diagram of a peak window provided by another embodiment of the present invention;

Fig. 3 is a schematic diagram of a peak window provided by another embodiment of the present invention;

Fig. 4 is a schematic diagram of displaying search results provided by another embodiment of the present invention;

5 is a structural diagram of an analysis system based on a keyword-based event evolution process provided by an embodiment of the present invention;

FIG. 6 is a structural diagram of an analysis system for a keyword-based event evolution process provided by another embodiment of the present invention.

Detailed ways

FIG. 1 is a flowchart of an analysis method for a keyword-based event evolution process provided by an embodiment of the present invention. According to the microblog information searched by keywords, the embodiment of the present invention counts the number of microblogs, and provides an analysis method for event evolution process based on keywords. The specific steps of the method are as follows:

Step S101, counting the number of search results in each time period, and obtaining a sequence in which the number of search results changes with the time period;

Before the statistics of the number of search results in each time period, it also includes:

Searching according to keywords and obtaining search results related to the keywords, the search results including time information;

The statistics of the number of search results in each time period include:

Count the number of search results in each time period according to the time information.

Through the keyword entered by the user, search and obtain the microblog information related to the keyword, specifically the microblog information containing the keyword, the microblog information specifically includes the microblog content and release time, and the release time is counted in a predetermined time period The number of microblog information in this time period, taking the time period as 1 day as an example, count the number of microblog information published in a certain day and contain this keyword, and count each time period in multiple time periods The number of microblog information in the time period and the number of microblog information correspond one-to-one, and the number of time periods and the number of microblog information constitutes a sequence in which the number of microblog information changes with time periods. Multiple time periods do not overlap each other.

Step S102, performing peak detection on the sequence to obtain at least one peak window, each of the peak windows includes a subsequence, the subsequence includes a spike, and the spike is the number of search results in the sequence local maximum;

The number of microblog information obtained in the above steps constitutes a sequence of changes in the number of microblog information with time periods for peak detection, and detects the local maximum value of the number of microblog information in the sequence. The maximum represents a spike, and the local maximum of the sequence includes at least one, that is, a sequence includes at least one spike. All the peaks of the sequence are acquired through peak detection, and the acquisition of the peaks is specifically obtained through the peak window, and the peak window is a subsequence including the peak.

Step S103, performing text analysis on the search results corresponding to the peak window to obtain a peak description;

Perform text analysis on all microblog information in the peak window, obtain the word segmentation in the text, use the word segmentation with high frequency in the text as the peak description, and explain the reason for the spike.

Step S104, displaying the sequence of the number of search results changing with the time period, and displaying the description of the peak at the position of the peak.

Display the sequence of changes in the number of microblog information over time, and display the peak description at the peak position, so that users can see at a glance the change in the number of microblog information related to the input keyword and the development trend of the situation related to the keyword , and public attention to the situation.

The embodiment of the present invention obtains the local maximum of the sequence by counting the number of search results changing with the time period according to the peak detection, analyzes the cause of the occurrence of the local maximum to obtain the peak description, and displays the sequence of the number of search results changing with the time period, And the peak description is displayed at the peak position, and the microblog information display results can show the trend of the amount of microblog information changing over time, and the trend of the amount of microblog information changing over time can show the evolution process of events related to keywords. Peak description users can know the main content of the microblog information without reading the microblog content in detail, which increases the amount of information displayed in the search results.

FIG. 2 is a schematic diagram of a peak window provided by another embodiment of the present invention; FIG. 3 is a schematic diagram of a peak window provided by another embodiment of the present invention. On the basis of the above embodiments, the spike window includes a time period identifier for the start of the window, a time period identifier for the peak and a time period identifier for the end of the window;

The performing peak detection according to the time periods and the number of search results in each time period to obtain the peak window includes:

Initialization parameter mean=C ₁ , C ₁ represents the number of search results in the first time period;

like Then update the initialization parameter Wherein, _Ci represents the number of search results in the i-th time period, and n represents the total number of the time period;

like It is then determined that the time period at which the window starts is identified as i, if Then j=j+1, continue to judge is established until When it is determined that the time period at which the window ends is identified as j, calculate C _k , i≤k≤j, so that are all established, then k represents the time period identification of the peak;

The peak window is window(i,k,j);

Update the initialization parameters Continue to acquire the peak window of the sequence according to the method of acquiring window(i,k,j).

As shown in Figure 2, if C ₄ is significantly larger, the initialization parameter The corresponding increase makes the subsequent peak window of the sequence difficult to be detected. In order to solve this problem, the embodiment of the present invention further includes: arranging in reverse order the sequence in which the number of search results changes with the time period to obtain a reverse sequence;

Acquiring the spike window of the reverse sequence according to the method of acquiring window(i,k,j);

The spike windows of the sequence and the spike windows of the reverse sequence are combined into a spike window set, and the repeated spike windows are recorded once.

Arrange the sequence in Figure 2 in reverse order to obtain the reverse sequence, and obtain the peak window of the reverse sequence according to the method of obtaining window(3,4,6). Specifically, the peak window can be detected along the reverse direction of the X axis. The specific detection method is the same as the above steps Similarly, the first detected peak window is window(8,10,12), since there is no significantly larger value in window(8,10,12), the initialization parameters It will not increase significantly, and the second detection window (3,4,6) can also be detected.

For the sequence in Figure 2, window (3,4,6) is obtained by forward detection along the X axis, and window (8,10,12) and window (3,4,6) are obtained by reverse detection along the X axis. Spike windows detected in two directions are merged into a set of spike windows {window(3,4,6), window(8,10,12)}, where the repeated spike window window(3,4,6) is recorded once, such as As shown in FIG. 3 , the set of spike windows is used as the final spike window of the sequence.

The embodiment of the present invention provides a specific method for obtaining the peak window, and through forward and reverse detection of the peak window of the sequence, the probability of the peak being detected in the sequence is greatly improved, and the calculation accuracy is improved.

Fig. 4 is a schematic diagram of displaying search results provided by another embodiment of the present invention. On the basis of the above embodiments, performing text analysis on the search results corresponding to the spike window, and obtaining the spike description includes:

Obtain the search result corresponding to the peak window, use the word segmentation tool to obtain the corresponding word segmentation of the search result; calculate the word frequency reverse document frequency (Term Frequency Inverse Document Frequency, referred to as TFIDF) value of each word segmentation; if the TFIDF value is greater than the threshold value, The word segmentation corresponding to the TFIDF value is used as the peak description.

Perform text analysis on all microblog information in the spike window, use the word segmentation tool to obtain the word segmentation in the text, and calculate the TFIDF value of each word segmentation, the definition of the TFIDF value of the hth word segmentation is as follows formula (1)(2)(3) :

TFIDF _h =TF _h *IDF _h (1)

TF _h ＝ N _h,t /Σ _m N _m,t (2)

IDF _h = 1/DF _h (3)

Among them, N _h,t represents the number of times that the h-th word segment appears in the t-th peak window, Σ _m N _m,t represents the total number of word segments that appear in the microblog information corresponding to the t-th peak window, and DF _h represents the number of words that appear in the t-th peak window The number of microblog information with h participle.

The word segmentation corresponding to the TFIDF value greater than the threshold is used as the peak description. Specifically, the TFIDF values are sorted in descending order, and the word segmentation corresponding to the first 5 TFIDF values is used as the peak description. Participles are the five most frequently occurring participle in Weibo information, and can summarize the main content of the event.

The displaying the sequence of the number of search results varying with the time period includes: connecting the sequence of the number of search results varying with the time period into a curve, and displaying the curve.

Taking the Malaysia Airlines accident as an example, as shown in Figure 4, the horizontal axis represents time, and the vertical axis represents the number of original Weibo. Original Weibo includes many reports on the Malaysia Airlines accident. One day is used as the time period for counting the number of original Weibo. , specifically from March 1, 2014 to July 31, 2014, counting the number of original microblogs about the Malaysia Airlines accident every other day or two days (24 hours a day), and obtaining the number of original microblogs over time Change the sequence, and connect the sequence into a curve, and obtain all the peak windows of the sequence according to the method of the above-mentioned embodiment, and the peak description corresponding to each peak window, and display the peak description at the peak position, as shown in Figure 4 As shown, the sequence includes two spikes. The first spike is described as Malaysia Airlines, MH370, lost contact, plane, blessing, and the second spike is described as Malaysia Airlines, MH17, Ukraine, shot down, and crashed.

In the embodiment of the present invention, by inputting a keyword, a continuous graph of the number of original microblogs containing the keyword changing over time is obtained. The continuous graph can show the evolution process of events related to the keyword, and at the peak of the continuous curve Display peak descriptions that can summarize Weibo content, increasing the amount of information displayed in search results.

FIG. 5 is a structural diagram of an analysis system for a keyword-based event evolution process provided by an embodiment of the present invention. The analysis system based on the keyword-based event evolution process provided by the embodiment of the present invention can execute the processing flow provided by the embodiment of the keyword-based event evolution process analysis method embodiment, as shown in FIG. 5 , the analysis of the keyword-based event evolution process The system 50 includes a statistical module 51, a detection module 52, a text analysis module 53 and a display module 54, wherein the statistical module 51 is used to count the number of search results in each time period, and obtain the number of search results that changes with the time period. sequence; the detection module 52 is used to perform peak detection on the sequence to obtain at least one peak window, and each of the peak windows includes a subsequence, and the subsequence includes a peak, and the peak is that the number of search results is within the specified peak. The local maximum in the sequence; the text analysis module 53 is used to perform text analysis on the search results corresponding to the peak window to obtain a peak description; the display module 54 is used to display the sequence that the number of search results changes with the time period , and display the spike description at the spike's location.

The embodiment of the present invention obtains the local maximum of the sequence by counting the number of search results changing with the time period according to the peak detection, analyzes the cause of the occurrence of the local maximum to obtain the peak description, and displays the sequence of the number of search results changing with the time period, And the peak description is displayed at the peak position, and the microblog information display results can show the trend of the amount of microblog information changing over time, and the trend of the amount of microblog information changing over time can show the evolution process of events related to keywords. Peak description users can know the main content of the microblog information without reading the microblog content in detail, which increases the amount of information displayed in the search results.

FIG. 6 is a structural diagram of an analysis system for a keyword-based event evolution process provided by another embodiment of the present invention. On the basis of Fig. 5, the peak window includes the time period identification of the window start, the time period identification of the peak and the time period identification of the end of the window; the detection module 52 is specifically used to initialize the parameter mean=C ₁ , C ₁ represents the first The number of search results in a time period; if Then update the initialization parameter Wherein, _Ci represents the number of search results in the i-th time period, and n represents the total number of the time period; if It is then determined that the time period at which the window starts is identified as i, if Then j=j+1, continue to judge is established until When it is determined that the time period at which the window ends is identified as j, calculate C _k , i≤k≤j, so that All established, then k represents the time period identification of the peak; the peak window is window (i, k, j); update the initialization parameters i=j+1, continue to acquire the peak window of the sequence according to the method of acquiring window(i,k,j).

The analysis system 50 of the event evolution process based on keywords also includes a reverse sequence module 55, and the reverse sequence module 55 is used to reversely sequence the sequence of the number of search results changing with the time period to obtain a reverse sequence; the detection module 52 is also used to Acquiring the spike window of the reverse sequence according to the method of obtaining window(i, k, j); merging the spike window of the sequence and the spike window of the reverse sequence into a set of spike windows, repeated The spike window is recorded once.

The analysis system 50 of the event evolution process based on keywords also includes a search module 49, and the search module 49 is used to search according to keywords and obtain search results relevant to the keywords, and the search results include time information; the statistics module 51 specifically It is used to separately count the number of search results in each time period according to the time information.

The text analysis module 53 is specifically used to obtain the search result corresponding to the peak window, and utilizes the word segmentation tool to obtain the word segmentation corresponding to the search result; calculate the word frequency reverse file frequency TFIDF value of each word segmentation; if the TFIDF value is greater than the threshold value, the The word segmentation corresponding to the TFIDF value is used as the peak description.

The display module 54 is specifically configured to connect the series of changes in the number of search results with the time period into a curve, and display the curve.

The keyword-based event evolution process analysis system provided by the embodiment of the present invention can be specifically used to execute the method embodiment provided in FIG. 1 above, and the specific functions will not be repeated here.

The embodiment of the present invention provides a specific method for obtaining the peak window, and through the forward and reverse detection of the peak window of the sequence, the probability of the peak being detected in the sequence is greatly improved, and the calculation accuracy is improved; The continuous curve graph of the number of original microblogs of keywords changing over time, the continuous curve graph can show the evolution process of events related to keywords, and the peak description of the microblog content can be displayed at the peak of the continuous curve, which improves the The amount of information displayed in the search results.

To sum up, the embodiment of the present invention obtains the local maximum value of the sequence by counting the sequence in which the number of search results changes over time according to peak detection, analyzes the cause of the local maximum value to obtain a peak description, and shows that the number of search results changes with time. The sequence of segment changes, and display the peak description at the peak position, realize the microblog information display results can show the trend of the number of microblog information over time, and the trend of the number of microblog information over time can show events related to keywords According to the evolution process of the peak description, users can know the main content of the microblog information without reading the microblog content in detail, which improves the amount of information displayed in the search results; provides a specific method to obtain the peak window, and through the forward and reverse detection sequence The peak window in the sequence greatly increases the probability of the peak being detected in the sequence and improves the calculation accuracy; by inputting a keyword, a continuous curve graph of the number of original microblogs containing the keyword changing over time is obtained, and the continuous curve graph can show The evolution process of events related to keywords, and the peak description that can summarize the content of Weibo is displayed at the peak of the continuous curve, which improves the amount of information displayed in the search results.

In the several embodiments provided by the present invention, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components can be combined or May be integrated into another system, or some features may be ignored, or not implemented. In another point, the mutual coupling or direct coupling or communication connection shown or discussed may be through some interfaces, and the indirect coupling or communication connection of devices or units may be in electrical, mechanical or other forms.

The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place, or may be distributed to multiple network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the solution of this embodiment.

In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, each unit may exist separately physically, or two or more units may be integrated into one unit. The above-mentioned integrated units can be implemented in the form of hardware, or in the form of hardware plus software functional units.

The above-mentioned integrated units implemented in the form of software functional units may be stored in a computer-readable storage medium. The above-mentioned software functional units are stored in a storage medium, and include several instructions to make a computer device (which may be a personal computer, server, or network device, etc.) or a processor (processor) execute the methods described in various embodiments of the present invention. partial steps. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (Read-Only Memory, ROM), random access memory (Random Access Memory, RAM), magnetic disk or optical disk and other various media that can store program codes. .

Those skilled in the art can clearly understand that for the convenience and brevity of description, only the division of the above-mentioned functional modules is used as an example for illustration. The internal structure of the system is divided into different functional modules to complete all or part of the functions described above. For the specific working process of the device described above, reference may be made to the corresponding process in the foregoing method embodiments, and details are not repeated here.

Finally, it should be noted that: the above embodiments are only used to illustrate the technical solutions of the present invention, rather than limiting them; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that: It is still possible to modify the technical solutions described in the foregoing embodiments, or perform equivalent replacements for some or all of the technical features; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the technical solutions of the various embodiments of the present invention. scope.

Claims (12)

1. An analysis method based on a keyword-based event evolution process, characterized in that, comprising: Counting the number of search results in each time period, and obtaining the sequence of the number of search results changing with the time period; Performing peak detection on the sequence to obtain at least one peak window, each of the peak windows includes a subsequence, the subsequence includes a spike, and the spike is a local maximum of the number of search results in the sequence ; Perform text analysis on the search results corresponding to the peak window to obtain a peak description; A sequence of changes in the number of search results with the time period is displayed, and a description of the peak is displayed at the position of the peak. 2. The method according to claim 1, wherein the peak window comprises a window start time period identification, a peak time period identification and a window end time period identification; Performing peak detection according to each time period and the number of search results in each time period to obtain a peak window includes: Initialization parameter mean=C ₁ , C ₁ represents the number of search results in the first time period; like Then update the initialization parameter Wherein, _Ci represents the number of search results in the i-th time period, and n represents the total number of the time period; like It is then determined that the time period at which the window starts is identified as i, if Then j=j+1, continue to judge is established until When it is determined that the time period at which the window ends is identified as j, calculate C _k , i≤k≤j, so that are all established, then k represents the time period identification of the peak; The peak window is window(i,k,j); Update the initialization parameters Continue to acquire the peak window of the sequence according to the method of acquiring window(i,k,j). 3. The method according to claim 2, further comprising: arranging in reverse order the sequence in which the number of search results changes with the time period to obtain a reverse sequence; Acquiring the spike window of the reverse sequence according to the method of acquiring window(i,k,j); The spike windows of the sequence and the spike windows of the reverse sequence are combined into a spike window set, and the repeated spike windows are recorded once. 4. The method according to claim 1, characterized in that before said counting the number of search results in each time period, it also includes: Searching according to keywords and obtaining search results related to the keywords, the search results including time information; The statistics of the number of search results in each time period include: Count the number of search results in each time period according to the time information. 5. The method according to claim 4, wherein said performing text analysis on the search results corresponding to said peak window, and obtaining a peak description comprises: Obtain the search result corresponding to the peak window, and use the word segmentation tool to obtain the word segmentation corresponding to the search result; Calculate the word frequency reverse file frequency TFIDF value of each word; If the TFIDF value is greater than the threshold, the word segmentation corresponding to the TFIDF value is used as the peak description. 6. The method according to any one of claims 1-5, wherein the sequence of displaying the number of search results changing with the time period comprises: The sequence of the number of search results changing with the time period is connected into a curve, and the curve is displayed. 7. An analysis system based on a keyword-based event evolution process, characterized in that it comprises: A statistics module, configured to count the number of search results in each time period, and obtain a sequence in which the number of search results changes with the time period; A detection module, configured to perform peak detection on the sequence to obtain at least one peak window, each of the peak windows includes a subsequence, the subsequence includes a peak, and the peak is the number of search results in the sequence Local maximum in ; A text analysis module, configured to perform text analysis on the search results corresponding to the peak window to obtain a peak description; A display module, configured to display the sequence of the number of search results changing with the time period, and display the description of the peak at the position of the peak. 8. the analysis system based on the event evolution process of keyword according to claim 7, is characterized in that, described spike window comprises the time period identification of window beginning, the time period identification of peak and the time period identification of window end; The detection module is specifically used to initialize the parameter mean=C ₁ , where C ₁ represents the number of search results in the first time period; if Then update the initialization parameter Wherein, _Ci represents the number of search results in the i-th time period, and n represents the total number of the time period; if It is then determined that the time period at which the window starts is identified as i, if Then j=j+1, continue to judge is established until When it is determined that the time period at which the window ends is identified as j, calculate C _k , i≤k≤j, so that All established, then k represents the time period identification of the peak; the peak window is window (i, k, j); update the initialization parameters Continue to acquire the peak window of the sequence according to the method of acquiring window(i,k,j). 9. the analysis system based on the event evolution process of keyword according to claim 8, is characterized in that, also comprises: A reverse order module, configured to reversely arrange the sequence in which the number of search results changes with the time period to obtain a reverse sequence; The detection module is also used to obtain the spike window of the reverse sequence according to the method of obtaining window(i,k,j); merge the spike window of the sequence with the spike window of the reverse sequence For a set of spike windows, the spike windows that are repeated are recorded once. 10. the analysis system based on the event evolution process of keyword according to claim 7, is characterized in that, also comprises: A search module, configured to search according to keywords and obtain search results related to the keywords, the search results including time information; The statistical module is specifically configured to count the number of search results in each time period according to the time information. 11. The analysis system based on the keyword-based event evolution process according to claim 10, wherein the text analysis module is specifically used to obtain the search results corresponding to the spike window, and obtain the search results using word segmentation tools Corresponding participle; calculate the word frequency inverse document frequency TFIDF value of each participle; If described TFIDF value is greater than threshold value, the participle corresponding to described TFIDF value is described as described spike. 12. The analysis system based on the keyword-based event evolution process according to any one of claims 7-11, characterized in that, the display module is specifically used to change the sequence of the number of search results with the time period Connected to a curve, display the curve.