CN108595564B - Method and device for evaluating media friendliness and computer-readable storage medium - Google Patents
Method and device for evaluating media friendliness and computer-readable storage medium Download PDFInfo
- Publication number
- CN108595564B CN108595564B CN201810330401.1A CN201810330401A CN108595564B CN 108595564 B CN108595564 B CN 108595564B CN 201810330401 A CN201810330401 A CN 201810330401A CN 108595564 B CN108595564 B CN 108595564B
- Authority
- CN
- China
- Prior art keywords
- media
- friendliness
- emotion
- text
- texts
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 71
- 238000003860 storage Methods 0.000 title claims abstract description 12
- 230000008451 emotion Effects 0.000 claims abstract description 153
- 238000004458 analytical method Methods 0.000 claims abstract description 48
- 238000009499 grossing Methods 0.000 claims abstract description 33
- 238000011156 evaluation Methods 0.000 claims abstract description 23
- 230000007774 longterm Effects 0.000 claims abstract description 19
- 238000006243 chemical reaction Methods 0.000 claims abstract description 13
- 239000013598 vector Substances 0.000 claims description 31
- 238000004364 calculation method Methods 0.000 claims description 29
- 230000007935 neutral effect Effects 0.000 claims description 16
- 238000010801 machine learning Methods 0.000 claims description 8
- 238000004422 calculation algorithm Methods 0.000 claims description 7
- 238000010586 diagram Methods 0.000 description 10
- 230000002996 emotional effect Effects 0.000 description 10
- 230000008569 process Effects 0.000 description 9
- 230000006870 function Effects 0.000 description 8
- 238000004590 computer program Methods 0.000 description 7
- 238000012545 processing Methods 0.000 description 7
- 238000012544 monitoring process Methods 0.000 description 6
- 238000012986 modification Methods 0.000 description 5
- 230000004048 modification Effects 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 3
- 238000010606 normalization Methods 0.000 description 2
- 230000011218 segmentation Effects 0.000 description 2
- 238000005728 strengthening Methods 0.000 description 2
- 238000012935 Averaging Methods 0.000 description 1
- 238000012300 Sequence Analysis Methods 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000010276 construction Methods 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 238000009826 distribution Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 230000006855 networking Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 230000001105 regulatory effect Effects 0.000 description 1
- 230000033764 rhythmic process Effects 0.000 description 1
- 230000007480 spreading Effects 0.000 description 1
- 238000003892 spreading Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Images
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a method and a device for evaluating media friendliness and a computer-readable storage medium, which relate to the technical field of public opinion analysis, and the method comprises the following steps: acquiring a plurality of related texts which are issued by target media and related to a public sentiment object; performing emotion polarity analysis on each relevant text in the plurality of relevant texts to obtain an emotion polarity value of each relevant text; respectively converting the emotion polarity value of each relevant text into a media friendliness index corresponding to each relevant text, and forming a media friendliness index time sequence; smoothing the media friendliness index time sequence to obtain a smoothed media friendliness index time sequence; and evaluating the media friendliness of the target media based on the smoothed media friendliness index time sequence. The embodiment of the invention can realize the conversion from the text emotion to the media long-term attitude, thereby realizing the accurate evaluation of the long-term stable attitude of the target media to the public sentiment object.
Description
Technical Field
The invention relates to the technical field of public opinion analysis, in particular to a method and a device for evaluating media friendliness and a computer-readable storage medium.
Background
For some hot spots and focus problems in real life, traditional media, opinion leaders and the common people can issue pronounced opinions and opinions. The public opinion and opinion published for an event or phenomenon is called public opinion, the holding and publishing people of the public opinion and opinion are called public opinion subjects, the objects concerned and commented by the public opinion subjects are public opinion objects, and the media (including network media and traditional paper media) published by the public opinion and opinion is a public opinion carrier. Public opinion monitoring is the behavior of monitoring, analyzing and predicting the opinions and opinions published by public opinion subjects related to public opinion objects.
The media has the characteristics of professional collection and editing, public distribution, wide audience and the like, is a main source for people to obtain information, and has important influence on the audience, so the media is a main object for public opinion monitoring. With the continuous popularization of the internet, the network has become a basic tool for work and life, the online media has become a big trend, the influence of emerging network media is larger and larger, and most of the traditional paper media release network versions at the same time. The media networking speeds up the rhythm of media sending and spreading, and the mass appearance of network public sentiment provides new challenges and opportunities for public sentiment monitoring and analysis. From the technical perspective, the network public opinion monitoring refers to the integration of an internet information acquisition technology and an information intelligent processing technology, and realizes information requirements of network public opinion monitoring, news topic tracking and the like of a user by automatically capturing mass information of the internet, automatically classifying and clustering, topic detection and topic focusing, so as to form analysis results such as briefings, reports, charts and the like.
The public sentiment content not only conveys the objective opinion of a public sentiment subject on a public sentiment object, but also comprises the subjective feeling and emotion of the public sentiment subject. Text sentiment analysis, also known as opinion mining, is a process of analyzing, processing, generalizing and reasoning subjective text with sentiment colors. The initial emotion analysis was primarily directed to the analysis of words with emotional color, e.g., "nice" is a word with positive color and "ugly" is a word with depreciative color. Later, related researchers gradually transitioned from the analysis of simple emotional words to the study of more complex emotional sentences and emotional chapters; in general, the polarity of subjective text is divided into two categories, positive and negative, or three categories, positive, neutral and negative.
The text sentiment analysis technology is oriented to a single document, and the long-term attitude of a public sentiment object cannot be reflected by the sentiment tendency of the single document. The goal of media friendliness assessment is to observe the long-term stable attitude of media to public opinion objects (such as a company). Knowledge of media friendliness can be used to identify media of different tendencies, giving different degrees of attention.
However, the inventors found that the evaluation and tracking of media friendliness has not been disclosed in the prior art.
Disclosure of Invention
In order to solve the problems in the prior art, embodiments of the present invention provide a method and an apparatus for evaluating media friendliness, so as to evaluate the media friendliness of a media to a public sentiment object, thereby knowing its long-term stable attitude.
The embodiment of the invention provides the following specific technical scheme:
in a first aspect, a method for evaluating media friendliness is provided, the method comprising the steps of:
acquiring a plurality of related texts which are issued by target media and related to a public sentiment object;
performing emotion polarity analysis on each relevant text in the plurality of relevant texts to obtain an emotion polarity value of each relevant text;
respectively converting the emotion polarity values of the relevant texts into media friendliness indexes corresponding to the relevant texts, and forming a media friendliness index time sequence;
smoothing the media friendliness index time sequence to obtain the smoothed media friendliness index time sequence;
and evaluating the media friendliness of the target media based on the smoothed media friendliness index time sequence.
With reference to the first aspect, in a first possible implementation manner, the method further includes:
continuously acquiring the plurality of related texts;
and continuously evaluating and updating the media friendliness of the target media in a preset time window based on the plurality of continuously acquired related texts.
With reference to the first aspect or the first possible implementation manner of the first aspect, in a second possible implementation manner, the step of obtaining a plurality of relevant texts related to a public opinion object published by a target media further includes:
acquiring a media text set issued by the target media;
generating a first vector space model for each media text in the set of media texts; and
generating a second vector space model of the public opinion object;
calculating the similarity of each first vector space model and the second vector space model;
and if the similarity obtained by calculation exceeds a preset threshold value, determining that the media text is the related text related to the public sentiment object.
With reference to the first aspect, in a third possible implementation manner, the step of performing emotion polarity analysis on each of the plurality of relevant texts further includes:
and carrying out emotion polarity analysis on each relevant text based on emotion classification of an emotion dictionary or a machine learning algorithm.
With reference to the first aspect to any one of the third possible implementation manners of the first aspect, in a fourth possible implementation manner, the emotion polarities include a positive emotion, a neutral emotion, and a negative emotion, and the step of converting the emotion polarity values of the respective related texts into the media friendliness indexes corresponding to the respective related texts further includes:
respectively calculating to obtain media friendliness indexes corresponding to the relevant texts according to a predefined media friendliness index calculation formula and the emotion polarity values of the relevant texts;
wherein the predefined media friendliness index calculation formula adopts any one of the following formulas:
Fr＝p+x-n；
Fs＝p-x-n；
wherein, FrFor loose media friendliness index, FsFor strict media friendliness index, p represents the probability of positive emotion, x represents the probability of neutral emotion, n represents the probability of negative emotion, 0 ≦ p, x, n ≦ 1, p + x + n ≦ 1, -1 ≦ Fr,Fs≤1。
With reference to the first aspect, in a fifth possible implementation manner, before the step of forming a time series of media friendliness indexes, the method further includes:
carrying out average calculation on the media friendliness indexes corresponding to the relevant texts according to a time window to obtain an average value of the media friendliness indexes corresponding to the time window;
the step of forming a time series of media friendliness indices further comprises:
and forming the media friendliness index time sequence based on a plurality of average values corresponding to a plurality of time windows.
With reference to the first aspect or the first or fifth possible implementation manner of the first aspect, in a sixth possible implementation manner, the step of smoothing the media friendliness index time sequence to obtain the smoothed media friendliness index time sequence further includes:
and smoothing the media friendliness index time sequence by adopting a moving average method or a weighted moving average method.
With reference to the first aspect or the first possible implementation manner of the first aspect, in a seventh possible implementation manner, after the step of evaluating the media friendliness of the target media based on the smoothed media friendliness index time series, the method further includes:
and predicting the future media friendliness index of the target media according to the evaluation result of the media friendliness of the target media.
In a second aspect, there is provided a media friendliness assessment apparatus, the apparatus comprising:
the text acquisition module is used for acquiring a plurality of related texts which are issued by target media and related to the public sentiment object;
the emotion analysis module is used for carrying out emotion polarity analysis on each relevant text in the plurality of relevant texts to obtain an emotion polarity value of each relevant text;
the index conversion module is used for respectively converting the emotion polarity values of the related texts into media friendliness indexes corresponding to the related texts;
the sequence forming module is used for forming a media friendliness index time sequence;
the smoothing module is used for smoothing the media friendliness index time sequence to obtain the smoothed media friendliness index time sequence;
and the friendliness evaluation module is used for evaluating the media friendliness of the target media based on the smoothed media friendliness index time sequence.
With reference to the second aspect, in a first possible implementation manner, the text obtaining module is specifically configured to:
continuously acquiring the plurality of related texts;
the friendliness evaluation module is specifically configured to:
and continuously evaluating and updating the media friendliness of the target media in a preset time window based on the plurality of continuously acquired related texts.
With reference to the second aspect or the first possible implementation manner of the second aspect, in a second possible implementation manner, the text obtaining module is specifically configured to:
acquiring a media text set issued by the target media;
generating a first vector space model for each media text in the set of media texts; and
generating a second vector space model of the public opinion object;
calculating the similarity of each first vector space model and the second vector space model;
and if the similarity obtained by calculation exceeds a preset threshold value, determining that the media text is the related text related to the public sentiment object.
With reference to the second aspect, in a third possible implementation manner, the emotion analysis module is specifically configured to:
and carrying out emotion polarity analysis on each relevant text based on emotion classification of an emotion dictionary or a machine learning algorithm.
With reference to the second aspect to any one of the third possible implementation manners of the second aspect, in a fourth possible implementation manner, the emotion polarities include a positive emotion, a neutral emotion, and a negative emotion, and the index conversion module is specifically configured to:
respectively calculating to obtain media friendliness indexes corresponding to the relevant texts according to a predefined media friendliness index calculation formula and the emotion polarity values of the relevant texts;
wherein the predefined media friendliness index calculation formula adopts any one of the following formulas:
Fr＝p+x-n；
Fs＝p-x-n；
wherein, FrFor loose media friendliness index, FsFor strict media friendliness index, p represents the probability of positive emotion, x represents the probability of neutral emotion, n represents the probability of negative emotion, 0 ≦ p, x, n ≦ 1, p + x + n ≦ 1, -1 ≦ Fr,Fs≤1。
With reference to the second aspect, in a fifth possible implementation manner, the apparatus further includes:
the average calculation module is used for carrying out average calculation on the media friendliness indexes corresponding to the relevant texts according to a time window to obtain an average value of the media friendliness indexes corresponding to the time window;
the sequence forming module is further configured to form the media friendliness index time sequence based on a plurality of the averages corresponding to a plurality of the time windows.
With reference to the second aspect or the fifth possible implementation manner of the first or second aspect of the second aspect, in a sixth possible implementation manner, the smoothing module is specifically configured to:
and smoothing the media friendliness index time sequence by adopting a moving average method or a weighted moving average method.
With reference to the second aspect or the first possible implementation manner of the second aspect, in a seventh possible implementation manner, the apparatus further includes:
and the index prediction module is used for predicting the future media friendliness index of the target media according to the evaluation result of the media friendliness of the target media.
In a third aspect, a media friendliness assessment apparatus is provided, the apparatus comprising:
one or more processors;
a memory;
the program stored in the memory, when executed by the one or more processors, causes the processors to perform the steps of the media friendliness assessment method of any one of the above first aspects.
In a fourth aspect, there is provided a computer-readable storage medium storing a program which, when executed by a processor, causes the processor to perform the steps of the media-friendliness assessment method according to any one of the above first aspects.
The embodiment of the invention provides a method and a device for evaluating media friendliness and a computer readable storage medium, which are used for respectively converting emotion polarity values of related texts into media friendliness indexes corresponding to the related texts on the basis of emotion analysis of the related texts issued by a target medium, forming a media friendliness index time sequence, smoothing the media friendliness index time sequence, and evaluating the media friendliness of the target medium by using the smoothed media friendliness index time sequence, so that conversion from text emotion to media attitude is realized, and the long-term stable attitude of the target medium to a public opinion object can be accurately evaluated.
Drawings
In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings needed to be used in the description of the embodiments will be briefly introduced below, and it is obvious that the drawings in the following description are only some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.
FIG. 1 is a flowchart of a method for evaluating media friendliness according to an embodiment of the present invention;
FIG. 2 is a diagram illustrating the smoothing of a media friendliness index time series according to an embodiment of the present invention;
fig. 3 is a block diagram of a media friendliness assessment apparatus according to an embodiment of the present invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is obvious that the described embodiments are only a part of the embodiments of the present invention, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
The embodiment of the invention provides a media friendliness evaluation method, which is based on emotion analysis of each related text issued by a target media, converts emotion polarity values of each related text into media friendliness indexes corresponding to each related text, forms a media friendliness index time sequence, smoothes the media friendliness index time sequence, and evaluates the media friendliness of the target media by using the smoothed media friendliness index time sequence, thereby realizing conversion from text emotion to media long-term attitude, and realizing accurate evaluation of the long-term stable attitude of the target media to a public sentiment object. The method provided in this embodiment may be applied to various devices, such as a desktop computer, a personal computer, a mobile terminal, and a server, which is not limited in this respect.
Referring to fig. 1, an embodiment of the present invention provides a flow chart of a media friendliness assessment method, where the method includes the following steps:
and S10, acquiring a plurality of related texts related to the public sentiment object and issued by the target media.
And S20, performing emotion polarity analysis on each relevant text in the plurality of relevant texts to obtain an emotion polarity value of each relevant text.
And S30, converting the emotion polarity values of the related texts into media friendliness indexes corresponding to the related texts respectively, and forming a media friendliness index time sequence.
And S40, smoothing the media friendliness index time sequence to obtain a smoothed media friendliness index time sequence.
And S50, evaluating the media friendliness of the target media based on the smoothed media friendliness index time series.
The embodiment of the invention provides a media friendliness evaluation method, which is characterized in that on the basis of emotion analysis of each related text issued by a target media, emotion polarity values of the related texts are respectively converted into media friendliness indexes corresponding to the related texts, a media friendliness index time sequence is formed, the media friendliness index time sequence is subjected to smoothing treatment, and the media friendliness of the target media is evaluated by using the smoothed media friendliness index time sequence, so that conversion from text emotion to media attitude is realized, and the media friendliness of the target media to a public sentiment object is accurately evaluated.
In some embodiments of the invention, the method further comprises:
continuously acquiring a plurality of related texts;
and continuously evaluating and updating the media friendliness of the target media in a preset time window based on the continuously acquired plurality of related texts.
The preset time interval may be set in a user-defined manner according to the observed time scale, for example, the preset time interval is set to be one week, which is not limited in the present invention.
Specifically, based on a plurality of continuously acquired related texts, the emotion polarity value of each related text in the plurality of continuously acquired related texts is obtained through analysis and is correspondingly converted into a media friendliness index corresponding to each related text, a previously-formed media friendliness index time sequence is updated, the updated media friendliness index time sequence is subjected to smoothing processing, and then the media friendliness of the target media is continuously evaluated and updated based on the smoothed media friendliness index time sequence.
In the embodiment of the invention, the media friendliness of the target media is continuously evaluated and updated based on a plurality of relevant texts which are issued by the target media and are relevant to the public sentiment object, so that the conversion from the text emotion to the long-term attitude of the media is realized, and the long-term stable attitude of the target media to the public sentiment object is accurately evaluated.
In some embodiments of the invention, step S10 further comprises:
step S101, acquiring a media text set issued by a target media;
the method comprises the steps of collecting a media text set issued by a target media according to a time sequence, and performing text preprocessing on the media text set, such as Chinese word segmentation and part-of-speech tagging on the media text set by using a word segmentation tool.
Step S102, a first vector space model of each media text in the media text set is generated, and a second vector space model of the public sentiment object is generated.
The process of step S102 may include:
step S1021, the media document and the public sentiment object are respectively expressed by vectors.
Media document DjCan be represented as Dj(w1j,w2j,…,wnj) Where n is the number of words in the system, wijRepresents the keyword i in the document DjThe weight in (1).
The vector of the public opinion object query Q can be represented as Q (w)1k,w2k,…,wnk)，wikRepresenting the weight of the word i in the query Q.
Step S1022, calculating the weight in the each-dimensional space model by adopting the TF × IDF manner, that is, the keyword i in the document DjWeight w ofij＝TFij+IDFij. The specific calculation process is as follows:
TF weights: TF (term frequency) is the number of times a word appears in a document, weight wij＝TFijOr a normalized TF value.
Normalization of TF (Normalization): normalizing the TF values of all the index words in a document to between 0,1 can generally be done in one of the following ways:
i)，wtf＝TF/max(TF)
ii)，wtfa + (1-a) × TF/max (TF), where a is a regulatory factor and the empirical value a is 0.5 or 0.4.
Document frequency of word df (document frequency): the document length of the word in the whole document set, DF reflects the word distinction degree, and higher DF means that the word is more common, therefore, the lower the distinction degree is, the lower the weight is.
Inverse document frequency (Inverse DF, IDF): the reciprocal of DF is usually calculated using the following formula (N is the number of all documents in the document set):
calculating the weight: keyword i in document DjWeight w ofij＝TFij+IDFij。
Step S103, calculating the similarity of each first vector space model and each second vector space model.
The degree of relevance (i.e., similarity) of documents and query terms may be determined by the relative position of their respective vectors in the vector space. There are many similarity calculation functions, and the more common is the cosine function of the included angle between two vectors. Thus media document DjThe similarity value with the public opinion object query Q is obtained by the following formula:
and step S104, if the calculated similarity exceeds a preset threshold, determining that the media text is related to the public sentiment object.
Wherein, a threshold value of query relevance can be set, when the calculated similarity is higher than the threshold value, the document can be regarded as a related document, and then the emotional tendency of the document to the public sentiment object is further analyzed.
In the embodiment of the invention, since the same media usually releases various contents, and the public sentiment analysis only concerns the situation of the public sentiment object concerned, the media friendliness evaluation needs to select reports related to the public sentiment object from all the contents released by the media, and if the public sentiment object concerned can be represented by a group of keywords, the selection of related texts can be represented as a text retrieval problem based on the keywords. The method selects the text related to the public sentiment object based on the vector space model, can balance the difficulty and effect of the implementation process, and can understand that in practical application, the selection of the text related to the public sentiment object can be realized by other methods, such as a Boolean model, a vector space model, a probability model and the like.
In some embodiments of the invention, step S20 further comprises:
and performing tripolar emotion analysis on each related text, wherein the analysis results are p, x, n, p is more than or equal to 0, x, n is less than or equal to 1, and p + x + n is 1, wherein p expresses the probability of positive emotion, x expresses the probability of neutral emotion, and n expresses the probability of negative emotion.
In one embodiment, emotion polarity analysis can be performed on each relevant text based on the emotion classification of the emotion dictionary, and specifically, the process can include:
step S201, selecting and perfecting an emotion dictionary. The emotion dictionary comprises a plurality of emotion words, and each emotion word has a score value representing emotion polarity intensity.
Step S202, combining the emotion strengthening words and the emotion negative words to score the emotion words and phrases. The emotion strengthening words are words for modifying the degree of emotion words and are used for expressing the intensity of emotion; the emotion negative words are words for negating emotion vocabularies, such as: words of no, not, absolutely not, no, not, impossible, none, failing, etc.
Step S203, accumulating and counting the emotion scores of each polarity emotion in the document, and respectively calculating the probability of each polarity emotion appearing in the document.
In the embodiment of the invention, the three-stage emotion analysis is carried out on each relevant text through emotion classification based on the emotion dictionary to obtain the emotion polarity value of each relevant text, and the media friendliness is presumed through the emotion analysis result based on a single document, so that the reliability and the feasibility of an evaluation basis can be ensured.
In another embodiment, emotion polarity analysis may be performed on each relevant text based on a machine learning algorithm, and specifically, the process may include:
step S211, constructing features: and a bag-of-words model (n-gram), part of speech, emotional words, emotional phrases, emotional transposed words and the like are adopted as characteristics.
Step S212, machine learning: naive Bayes, Maximum Entropy (Maximum Entropy) and SVM are used as classifiers for training.
And step S213, performing emotion polarity analysis by combining the relevant texts of the classifier to obtain emotion polarity values of the relevant texts.
It should be noted that, although it is preferable to perform tripolar emotion analysis on each relevant text in this embodiment, it is to be understood that the method provided in the embodiment of the present invention is not limited thereto, and a person skilled in the art may also perform bipolar emotion analysis or multi-polarity emotion analysis on each relevant text.
In the embodiment of the invention, the three-pole emotion analysis is carried out on each relevant text based on the machine learning algorithm to obtain the emotion polarity value of each relevant text, and the media friendliness is presumed based on the emotion analysis result of a single document, so that the reliability and feasibility of the evaluation basis can be ensured.
In some embodiments of the invention, the emotion polarities include positive emotion, neutral emotion and negative emotion, and step S30 further includes:
step S301, respectively calculating to obtain media friendliness indexes corresponding to the relevant texts according to a predefined media friendliness index calculation formula and the emotion polarity values of the relevant texts;
wherein, the predefined media friendliness index calculation formula adopts any one of the following formulas:
Fr＝p+x-n；
Fs＝p-x-n；
wherein, FrFor loose media friendliness index, FsFor strict media friendliness index, p represents the probability of positive emotion, x represents the probability of neutral emotion, n represents the probability of negative emotion, 0 ≦ p, x, n ≦ 1, p + x + n ≦ 1, -1 ≦ Fr,Fs≤1。
The difference between the relaxed and strict friendship index definitions is the utilization of neutral emotions, neutral emotions being considered positive by relaxed friendship and negative by strict friendship.
Step S302, a media friendliness index time sequence is formed based on the media friendliness indexes corresponding to the relevant texts.
Wherein, each element of the media friendliness index time sequence represents the emotional tendency of one text issued by the media to the public sentiment object.
In the embodiment of the invention, because the emotion embodied by a single document can not reflect the long-term attitude of the media, the emotion polarity value of each related text is converted into the media friendliness index corresponding to each related text, the media friendliness index time sequence is formed, and the media friendliness index time sequence is utilized, so that the conversion from the document emotion to the long-term attitude of the media is realized, and further the evaluation and tracking of the target media for the media friendliness of the public opinion object can be realized.
In some embodiments of the invention, step S40 further comprises:
and smoothing the media friendliness index time sequence by adopting a moving average method or a weighted moving average method.
The process of smoothing the media friendliness index time sequence by using the moving average method may include:
setting the time sequence as y1,y2,…,yt(ii) a The simple moving average method has the calculation formula as follows:
wherein M ist-a moving average over a period t; n-number of terms of moving average
When N is large, the calculation amount can be greatly reduced by using a recurrence formula.
The weighted moving average method can solve this problem because the moving average does not take into account the differences in the contribution of different documents to media friendliness.
The process of smoothing the media friendliness index time sequence by using the weighted moving average may include:
the general expression for weighted moving average is:
wherein M istw-a moving average over a period t; w is ai-yt-i+1By weight of (a) represents the corresponding ytImportance in weighted averaging.
The weights are chosen empirically, with the general principle: the weight of the recent data is large, and the weight of the future data is small.
Wherein, the weight can use the time attenuation function f (n) ═ A + e-μnWhere a is ∈ Z, and a > 1, μ ∈ (0,1), N is 1,2, …, N, denotes the sequence number of the emotion time series.
In the embodiment of the invention, a scene of public opinion application is combined, news documents are classified into original documents and reprinted documents, the original documents are usually the first originations, the attitude of media to a public opinion object can be reflected most, the tendency of the attitude reflected by retransmission is weak, and in order to reflect the difference, the values of the parameter A of the attenuation function of the document weight are different, so that in practical application, the original document corresponding to A is 1, and the non-original document corresponding to A is 0.5. The parameter μmay initially take a random value within a range of values, and may be estimated based on historical data when the historical data has accumulated to a certain extent, where the fitting may be implemented by python or matlab codes.
In the embodiment of the invention, the smoothed media friendliness index time sequence is obtained by smoothing the media friendliness index time sequence, and the media friendliness of the target media is evaluated based on the smoothed media friendliness index time sequence, so that the evaluation result of the target media on the media friendliness of the object to be treated public sentiment can be more stable, and the long-term stable attitude of the target media on the object to be treated public sentiment is further obtained.
In some embodiments of the present invention, before constructing the media friendliness index time series in step S40, the method may further comprise:
carrying out average calculation on the media friendliness indexes corresponding to the relevant texts according to the time window to obtain the average value of the corresponding media friendliness indexes in the time window;
the step of constructing the media friendliness index time series in step S40 further includes:
and forming a media friendliness index time sequence based on a plurality of average values corresponding to a plurality of time windows.
The smooth result is related to the time scale, if the observed time scale needs to be further enlarged, the time scale in a certain range, such as week, month, quarter, and the like, is selected, the media friendliness indexes corresponding to each relevant text in the plurality of relevant texts in the time window are averaged to obtain the mean value of the media friendliness indexes in the time window, then the media friendliness index time sequence is formed, and the media friendliness index time sequence is smoothed, so that the media friendliness condition of the large time scale can be further analyzed.
In some embodiments of the present invention, after step S50, the method may further comprise:
and predicting the media friendliness index of the target media in the future according to the evaluation result of the media friendliness of the target media.
Wherein, the prediction can be performed by using a weighted moving average, and the prediction formula is:
namely, the weighted moving average of the t th period is used as the predicted value of the t +1 th period.
The embodiment of the invention is based on the emotional analysis of a plurality of related texts of the target media, the media friendliness indication time sequence is formed by analyzing the media friendliness index corresponding to a single text, and the long-term stability attitude of the media to the public sentiment object is evaluated and predicted based on the smoothing treatment of the time sequence analysis, so that the accuracy of the evaluation and prediction of the long-term stability attitude of the media to the public sentiment object can be ensured.
Referring to fig. 3, in an embodiment of the present invention, a media friendliness assessment apparatus is further provided, including:
the text acquisition module 31 is configured to acquire a plurality of relevant texts related to the public sentiment object and issued by the target media;
the emotion analysis module 32 is configured to perform emotion polarity analysis on each relevant text in the multiple relevant texts to obtain an emotion polarity value of each relevant text;
the index conversion module 33 is configured to convert the emotion polarity values of the relevant texts into media friendliness indexes corresponding to the relevant texts respectively;
a sequence construction module 34 for constructing a media friendliness index time sequence;
the smoothing module 35 is configured to smooth the media friendliness index time sequence to obtain a smoothed media friendliness index time sequence;
and a friendliness evaluation module 36 for evaluating the media friendliness of the target media based on the smoothed media friendliness index time series.
Further, the text obtaining module 31 is specifically configured to:
continuously acquiring a plurality of related texts;
the friendliness assessment module 36 is specifically configured to:
and continuously evaluating and updating the media friendliness of the target media in a preset time window based on the continuously acquired plurality of related texts.
Further, the text obtaining module 31 is specifically configured to:
acquiring a media text set issued by a target media;
generating a first vector space model of each media text in the media text set; and
generating a second vector space model of the public sentiment object;
calculating the similarity of each first vector space model and each second vector space model;
and if the calculated similarity exceeds a preset threshold, determining that the media text is the related text related to the public sentiment object.
Further, the emotion analysis module 32 is specifically configured to:
and carrying out emotion polarity analysis on each relevant text based on emotion classification of an emotion dictionary or a machine learning algorithm.
Further, the emotion polarities include positive emotion, neutral emotion, and negative emotion, and the index conversion module 33 is specifically configured to:
respectively calculating to obtain a media friendliness index corresponding to each relevant text according to a predefined media friendliness index calculation formula and the emotion polarity value of each relevant text;
wherein, the predefined media friendliness index calculation formula adopts any one of the following formulas:
Fr＝p+x-n；
Fs＝p-x-n；
wherein, FrFor loose media friendliness index, FsFor strict media friendliness index, p represents the probability of positive emotion, x represents the probability of neutral emotion, n represents the probability of negative emotion, 0 ≦ p, x, n ≦ 1, p + x + n ≦ 1, -1 ≦ Fr,Fs≤1。
Further, the apparatus further comprises:
the average calculation module 37 is configured to perform average calculation on the media friendliness indexes corresponding to the relevant texts according to the time window to obtain an average value of the corresponding media friendliness indexes in the time window;
the sequence forming module 34 is further configured to form a media friendliness index time sequence based on a plurality of mean values corresponding to a plurality of time windows.
Further, the smoothing module 35 is specifically configured to:
and smoothing the media friendliness index time sequence by adopting a moving average method or a weighted moving average method.
Further, the apparatus further comprises:
and the index prediction module 38 is used for predicting the media friendliness index of the target media in the future according to the evaluation result of the media friendliness of the target media.
The embodiment of the invention provides a media friendliness evaluation device, which is characterized in that on the basis of emotion analysis of each related text issued by a target media, emotion polarity values of the related texts are respectively converted into media friendliness indexes corresponding to the related texts, a media friendliness index time sequence is formed, the media friendliness index time sequence is subjected to smoothing treatment, and the media friendliness of the target media is evaluated by using the smoothed media friendliness index time sequence, so that conversion from text emotion to media long-term attitude is realized, and the long-term stable attitude of the target media to a public opinion object is accurately evaluated.
In addition, an embodiment of the present invention further provides a device for evaluating media friendliness, where the device includes:
one or more processors;
a memory;
a program stored in the memory, which when executed by the one or more processors, causes the processors to perform the steps of the media friendliness assessment method of any of the above embodiments.
Another embodiment of the present invention further provides a computer-readable storage medium storing a program, which, when executed by a processor, causes the processor to perform the steps of the media friendliness assessment method in any one of the above embodiments.
As will be appreciated by one of skill in the art, embodiments of the present invention may be provided as a method, apparatus, or computer program product. Accordingly, embodiments of the present invention may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, embodiments of the present invention may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and the like) having computer-usable program code embodied therein.
Embodiments of the present invention are described with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the invention. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
While preferred embodiments of the present invention have been described, additional variations and modifications in those embodiments may occur to those skilled in the art once they learn of the basic inventive concepts. Therefore, it is intended that the appended claims be interpreted as including preferred embodiments and all such alterations and modifications as fall within the scope of the embodiments of the invention.
It will be apparent to those skilled in the art that various changes and modifications may be made in the present invention without departing from the spirit and scope of the invention. Thus, if such modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include such modifications and variations.
Claims (18)
1. A method for assessing media friendliness, the method comprising the steps of:
acquiring a plurality of related texts which are issued by target media and related to a public sentiment object;
performing emotion polarity analysis on each relevant text in the plurality of relevant texts to obtain an emotion polarity value of each relevant text;
respectively converting the emotion polarity values of the relevant texts into media friendliness indexes corresponding to the relevant texts, and forming a media friendliness index time sequence;
smoothing the media friendliness index time sequence to obtain the smoothed media friendliness index time sequence;
and evaluating the media friendliness of the target media to the public opinion object based on the smoothed media friendliness index time sequence, wherein the media friendliness is used for reflecting the long-term attitude of the target media to the public opinion object.
2. The method of claim 1, further comprising:
and continuously evaluating and updating the media friendliness of the target media in a preset time window based on the plurality of continuously acquired related texts.
3. The method of claim 1, wherein the step of obtaining a plurality of relevant texts related to the public opinion object published by the target media further comprises:
acquiring a media text set issued by the target media;
generating a first vector space model for each media text in the set of media texts; and
generating a second vector space model of the public opinion object;
calculating the similarity of each first vector space model and the second vector space model;
and if the similarity obtained by calculation exceeds a preset threshold value, determining that the media text is the related text related to the public sentiment object.
4. The method of claim 1, wherein the step of performing sentiment polarity analysis on each of the plurality of relevant texts further comprises:
and carrying out emotion polarity analysis on each relevant text based on emotion classification of an emotion dictionary or a machine learning algorithm.
5. The method according to any one of claims 1 to 4, wherein the emotion polarities include positive emotion, neutral emotion and negative emotion, and the step of converting the emotion polarity value of each relevant text into the media friendliness index corresponding to each relevant text further comprises:
respectively calculating to obtain media friendliness indexes corresponding to the relevant texts according to a predefined media friendliness index calculation formula and the emotion polarity values of the relevant texts;
wherein the predefined media friendliness index calculation formula adopts any one of the following formulas:
Fr＝p+x-n；
Fs＝p-x-n；
wherein, FrFor loose media friendliness index, FsFor strict media friendliness index, p represents the probability of positive emotion, x represents the probability of neutral emotion, n represents the probability of negative emotion, 0 ≦ p, x, n ≦ 1, p + x + n ≦ 1, -1 ≦ Fr,Fs≤1。
6. The method of claim 1, wherein the step of forming a time series of media friendliness indices is preceded by the method further comprising:
carrying out average calculation on the media friendliness indexes corresponding to the relevant texts according to a time window to obtain an average value of the media friendliness indexes corresponding to the time window;
the step of forming a time series of media friendliness indices further comprises:
and forming the media friendliness index time sequence based on a plurality of average values corresponding to a plurality of time windows.
7. The method according to claim 1 or 2, wherein the step of smoothing the media friendliness index time series to obtain the smoothed media friendliness index time series further comprises:
and smoothing the media friendliness index time sequence by adopting a moving average method or a weighted moving average method.
8. The method of claim 1 or 2, wherein after the step of evaluating the media friendliness of the target media based on the smoothed media friendliness index time series, the method further comprises:
and predicting the future media friendliness index of the target media according to the evaluation result of the media friendliness of the target media.
9. A media friendliness assessment apparatus, the apparatus comprising:
the text acquisition module is used for acquiring a plurality of related texts which are issued by target media and related to the public sentiment object;
the emotion analysis module is used for carrying out emotion polarity analysis on each relevant text in the plurality of relevant texts to obtain an emotion polarity value of each relevant text;
the index conversion module is used for respectively converting the emotion polarity values of the related texts into media friendliness indexes corresponding to the related texts;
the sequence forming module is used for forming a media friendliness index time sequence;
the smoothing module is used for smoothing the media friendliness index time sequence to obtain the smoothed media friendliness index time sequence;
and the friendliness evaluation module is used for evaluating the media friendliness of the target media to the public sentiment object based on the smoothed media friendliness index time sequence, wherein the media friendliness is used for reflecting the long-term attitude of the target media to the public sentiment object.
10. The apparatus of claim 9,
the text acquisition module is specifically configured to:
continuously acquiring the plurality of related texts;
the friendliness evaluation module is specifically configured to:
and continuously evaluating and updating the media friendliness of the target media in a preset time window based on the plurality of continuously acquired related texts.
11. The apparatus of claim 9, wherein the text acquisition module is specifically configured to:
acquiring a media text set issued by the target media;
generating a first vector space model for each media text in the set of media texts; and
generating a second vector space model of the public opinion object;
calculating the similarity of each first vector space model and the second vector space model;
and if the similarity obtained by calculation exceeds a preset threshold value, determining that the media text is the related text related to the public sentiment object.
12. The apparatus of claim 9, wherein the emotion analysis module is specifically configured to:
and carrying out emotion polarity analysis on each relevant text based on emotion classification of an emotion dictionary or a machine learning algorithm.
13. The apparatus of any of claims 9 to 12, wherein the emotion polarities include positive emotion, neutral emotion, and negative emotion, and wherein the index conversion module is specifically configured to:
respectively calculating to obtain media friendliness indexes corresponding to the relevant texts according to a predefined media friendliness index calculation formula and the emotion polarity values of the relevant texts;
wherein the predefined media friendliness index calculation formula adopts any one of the following formulas:
Fr＝p+x-n；
Fs＝p-x-n；
wherein, FrFor loose media friendliness index, FsFor strict media friendliness index, p represents the probability of positive emotion, x represents the probability of neutral emotion, and n representsProbability of negative emotion, 0 ≦ p, x, n ≦ 1, p + x + n ≦ 1, -1 ≦ Fr,Fs≤1。
14. The apparatus of claim 9, further comprising:
the average calculation module is used for carrying out average calculation on the media friendliness indexes corresponding to the relevant texts according to a time window to obtain an average value of the media friendliness indexes corresponding to the time window;
the sequence forming module is further configured to form the media friendliness index time sequence based on a plurality of the averages corresponding to a plurality of the time windows.
15. The apparatus according to claim 9 or 10, wherein the smoothing module is specifically configured to:
and smoothing the media friendliness index time sequence by adopting a moving average method or a weighted moving average method.
16. The apparatus of claim 9, further comprising:
and the index prediction module is used for predicting the future media friendliness index of the target media according to the evaluation result of the media friendliness of the target media.
17. A media friendliness assessment apparatus, the apparatus comprising:
one or more processors;
a memory;
the program stored in the memory, which when executed by the one or more processors, causes the processors to perform the steps of the media friendliness assessment method of any one of claims 1-8.
18. A computer-readable storage medium, characterized in that the computer-readable storage medium stores a program which, when executed by a processor, causes the processor to execute the steps of the media-friendliness assessment method according to any one of claims 1 to 8.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810330401.1A CN108595564B (en) | 2018-04-13 | 2018-04-13 | Method and device for evaluating media friendliness and computer-readable storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810330401.1A CN108595564B (en) | 2018-04-13 | 2018-04-13 | Method and device for evaluating media friendliness and computer-readable storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108595564A CN108595564A (en) | 2018-09-28 |
| CN108595564B true CN108595564B (en) | 2020-08-11 |
Family
ID=63622360
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810330401.1A Active CN108595564B (en) | 2018-04-13 | 2018-04-13 | Method and device for evaluating media friendliness and computer-readable storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108595564B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113282754A (en) * | 2021-06-10 | 2021-08-20 | 北京中科闻歌科技股份有限公司 | Public opinion detection method, device, equipment and storage medium for news events |
Family Cites Families (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102073646A (en) * | 2009-11-23 | 2011-05-25 | 北京科技大学 | Blog group-oriented subject propensity processing method and system |
| CN102789449B (en) * | 2011-05-20 | 2016-08-03 | 日电（中国）有限公司 | The method and apparatus that comment text is evaluated |
| WO2016035072A2 (en) * | 2014-09-02 | 2016-03-10 | Feelter Sales Tools Ltd | Sentiment rating system and method |
| CN104951548B (en) * | 2015-06-24 | 2018-04-20 | 烟台中科网络技术研究所 | A kind of computational methods and system of negative public sentiment index |
| CN106951409A (en) * | 2017-03-17 | 2017-07-14 | 黄淮学院 | A kind of network social intercourse media viewpoint tendency analysis system and method |
| CN107169632A (en) * | 2017-04-19 | 2017-09-15 | 广东数相智能科技有限公司 | Global media community image analysis method, device and system |
-
2018
- 2018-04-13 CN CN201810330401.1A patent/CN108595564B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN108595564A (en) | 2018-09-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11620455B2 (en) | Intelligently summarizing and presenting textual responses with machine learning | |
| CN106649818B (en) | Application search intention identification method and device, application search method and server | |
| CN106328147B (en) | Speech recognition method and device | |
| Desai et al. | Techniques for sentiment analysis of Twitter data: A comprehensive survey | |
| CN108829822B (en) | Media content recommendation method and device, storage medium and electronic device | |
| Godin et al. | Using topic models for twitter hashtag recommendation | |
| CN112733042B (en) | Recommendation information generation method, related device and computer program product | |
| CN111310476B (en) | Public opinion monitoring method and system using aspect-based emotion analysis method | |
| US9189470B2 (en) | Generation of explanatory summaries | |
| CN111930792B (en) | Labeling method and device for data resources, storage medium and electronic equipment | |
| CN104111925B (en) | Item recommendation method and device | |
| CN110263854B (en) | Live broadcast label determining method, device and storage medium | |
| CN109933782B (en) | User emotion prediction method and device | |
| WO2015021937A1 (en) | Method and device for user recommendation | |
| CN112329824A (en) | Multi-model fusion training method, text classification method and device | |
| CN110083837B (en) | Keyword generation method and device | |
| CN109784368A (en) | A kind of determination method and apparatus of application program classification | |
| CN113627151B (en) | Cross-modal data matching method, device, equipment and medium | |
| Wu et al. | ECNU at SemEval-2017 task 3: Using traditional and deep learning methods to address community question answering task | |
| CN111832305A (en) | User intention identification method, device, server and medium | |
| CN107092679B (en) | Feature word vector obtaining method and text classification method and device | |
| CN117351336A (en) | Image auditing method and related equipment | |
| Sabeeh et al. | Enhancing the fake news detection by applying effective feature selection based on semantic sources | |
| CN107908649B (en) | Text classification control method | |
| CN108595564B (en) | Method and device for evaluating media friendliness and computer-readable storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20240306 Address after: Room 1179, W Zone, 11th Floor, Building 1, No. 158 Shuanglian Road, Qingpu District, Shanghai, 201702 Patentee after: Shanghai Zhongan Information Technology Service Co.,Ltd. Country or region after: China Address before: 518000 Room 201, building A, No. 1, Qian Wan Road, Qianhai Shenzhen Hong Kong cooperation zone, Shenzhen, Guangdong (Shenzhen Qianhai business secretary Co., Ltd.) Patentee before: ZHONGAN INFORMATION TECHNOLOGY SERVICE Co.,Ltd. Country or region before: China |
|
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20240415 Address after: Room 1179, W Zone, 11th Floor, Building 1, No. 158 Shuanglian Road, Qingpu District, Shanghai, 201702 Patentee after: Shanghai Zhongan Information Technology Service Co.,Ltd. Country or region after: China Address before: 518000 Room 201, building A, No. 1, Qian Wan Road, Qianhai Shenzhen Hong Kong cooperation zone, Shenzhen, Guangdong (Shenzhen Qianhai business secretary Co., Ltd.) Patentee before: ZHONGAN INFORMATION TECHNOLOGY SERVICE Co.,Ltd. Country or region before: China |
|
| TR01 | Transfer of patent right |