CN107729337A - The monitoring method and device of event - Google Patents
The monitoring method and device of event Download PDFInfo
- Publication number
- CN107729337A CN107729337A CN201610659660.XA CN201610659660A CN107729337A CN 107729337 A CN107729337 A CN 107729337A CN 201610659660 A CN201610659660 A CN 201610659660A CN 107729337 A CN107729337 A CN 107729337A
- Authority
- CN
- China
- Prior art keywords
- event
- clause
- keyword
- events
- candidate
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/951—Indexing; Web crawling techniques
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a kind of monitoring method of event and device.Wherein, this method includes：The event keyword and affair character information of historical events are obtained, wherein, affair character information is used for the dimensional information for characterizing event keyword；The event keyword of historical events is trained, obtains the clause rule for monitoring online data；According to clause rules detection online data, candidate events are obtained；Screened according to affair character information from candidate events and obtain object event.The present invention is solved in the prior art, and accident is monitored by matching the keyword pre-set, causes the low technical problem of accident monitoring accuracy.
Description
Technical field
The present invention relates to application of net field, in particular to the monitoring method and device of a kind of event.
Background technology
In category is extracted in natural language processing and Web information, event excavates the backtracking being broadly divided into for historical events
With two aspects of monitoring for accident, because online data are caused in real time, therefore monitoring process can only rely on
In data in advance, and data afterwards can not be depended on, therefore, no matter from difficulty or in value, accident monitoring
All recall far above historical data.
In the monitoring process of accident, the key issue that determines monitoring result is the promptness of monitoring, more early monitoring
It is bigger to the generation of accident, the value of monitoring result.In addition, another key issue of accident monitoring is monitoring knot
The accuracy of fruit, because the response for accident needs certain cost, it is therefore desirable to ensure the standard of monitoring result as far as possible
True property.It is a kind of example in the case where not considering the extreme case of accuracy to return to whole Monitoring Datas in real time, and this, which is undoubtedly, meets
Promptness, but accuracy approximation rate zero.Therefore, accident monitoring is a synthesis for needing to take into account promptness and accuracy
Problem.
Generally, the method for handling accident monitoring is to pass through Keywords matching.The pass for needing to monitor is set in advance
Keyword list.When real time data arrives, check whether data include the word in lists of keywords, if comprising, then it is assumed that should
Data reflect accident.
But still have weak point in terms of accuracy above by the monitoring method of Keywords matching, use the above method
When being monitored, the dependent event of keyword is set to be extracted in advance, but be not provided with advance in keyword
Event can not then be identified, and so as to cause monitoring result not comprehensive, i.e. monitoring result accuracy is low.
For in the prior art, monitoring accident by matching the keyword pre-set, causing accident to be supervised
The problem of accuracy is low is surveyed, not yet proposes effective solution at present.
The content of the invention
The embodiments of the invention provide a kind of monitoring method of event and device, at least to solve in the prior art, to pass through
The keyword that pre-sets is matched to monitor accident, causes the low technical problem of accident monitoring accuracy.
One side according to embodiments of the present invention, there is provided a kind of monitoring method of event, including：Obtain historical events
Event keyword and affair character information, wherein, affair character information is used to characterize the dimensional information of event keyword；Training
The event keyword of historical events, obtain the clause rule for monitoring online data；According to clause rules detection online data,
Obtain candidate events；Screened according to affair character information from candidate events and obtain object event.
Another aspect according to embodiments of the present invention, a kind of monitoring device of event is additionally provided, including：First obtains mould
Block, for obtaining the event keyword and affair character information of historical events, wherein, affair character information is used to characterize event pass
The dimensional information of keyword；Training module, for training the event keyword of historical events, obtain the sentence for monitoring online data
Formula rule；Monitoring modular, for according to clause rules detection online data, obtaining candidate events；Screening module, for according to thing
Part characteristic information screens from candidate events and obtains object event.
Herein it should be noted that by artificial other means such as arrangement or log recording in advance, it can sum up and go through
The event keyword occurred in history, still, event (such as：Accident) not merely include what is had occurred and that in history
Event, in addition to some new events for not occurred in history.Therefore, it is difficult if be only monitored by event keyword
To monitor whole events.The scheme that the above embodiments of the present application provide is identical with the reporting style of historical events using new events
The characteristics of, the description of extensive specific event keyword, the clause rule obtained using the keyword of training historical events come to
Line number is according to being monitored, so as to ensure comprehensive, the accuracy of monitoring result.
The above embodiments of the present application obtain the event keyword and affair character information of historical events, train historical events
Event keyword, the clause rule for monitoring online data is obtained, according to clause rules detection online data, obtains candidate's thing
Part, screened according to affair character information from candidate events and obtain object event.Such scheme can ensure to monitor promptness
On the basis of, effectively improve it is above-mentioned for new accident None- identified the problem of, solve in the prior art, it is pre- by matching
The keyword that first sets monitors accident, causes the low technical problem of accident monitoring accuracy.
Brief description of the drawings
Accompanying drawing described herein is used for providing a further understanding of the present invention, forms the part of the application, this hair
Bright schematic description and description is used to explain the present invention, does not form inappropriate limitation of the present invention.In the accompanying drawings：
Fig. 1 is a kind of flow chart of the monitoring method of according to embodiments of the present invention 1 event；
Fig. 2 is a kind of schematic diagram of the monitoring device of according to embodiments of the present invention 2 event；
Fig. 3 is a kind of schematic diagram of the monitoring device of according to embodiments of the present invention 2 optional event；
Fig. 4 is a kind of schematic diagram of the monitoring device of according to embodiments of the present invention 2 optional event；
Fig. 5 is a kind of schematic diagram of the monitoring device of according to embodiments of the present invention 2 optional event；And
Fig. 6 is a kind of schematic diagram of the monitoring device of according to embodiments of the present invention 2 optional event.
Embodiment
In order that those skilled in the art more fully understand the present invention program, below in conjunction with the embodiment of the present invention
Accompanying drawing, the technical scheme in the embodiment of the present invention is clearly and completely described, it is clear that described embodiment is only
The embodiment of a part of the invention, rather than whole embodiments.Based on the embodiment in the present invention, ordinary skill people
The every other embodiment that member is obtained under the premise of creative work is not made, it should all belong to the model that the present invention protects
Enclose.
It should be noted that term " first " in description and claims of this specification and above-mentioned accompanying drawing, "
Two " etc. be for distinguishing similar object, without for describing specific order or precedence.It should be appreciated that so use
Data can exchange in the appropriate case, so as to embodiments of the invention described herein can with except illustrating herein or
Order beyond those of description is implemented.In addition, term " comprising " and " having " and their any deformation, it is intended that cover
Cover it is non-exclusive include, be not necessarily limited to for example, containing the process of series of steps or unit, method, system, product or equipment
Those steps or unit clearly listed, but may include not list clearly or for these processes, method, product
Or the intrinsic other steps of equipment or unit.
Embodiment 1
According to embodiments of the present invention, there is provided a kind of embodiment of the method for the monitoring method of event, it is necessary to explanation,
The step of flow of accompanying drawing illustrates can perform in the computer system of such as one group computer executable instructions, also,
, in some cases, can be with different from shown in order execution herein although showing logical order in flow charts
The step of going out or describing.
Fig. 1 is a kind of flow chart of the monitoring method of event according to embodiments of the present invention, as shown in figure 1, this method bag
Include following steps：
Step S102, the event keyword and affair character information of historical events are obtained, wherein, affair character information is used for
The dimensional information of sign event keyword.
Specifically, in above-mentioned steps, above-mentioned historical events is used to represent the event occurred before current time.
The basic logic of data monitoring is the result that will match to by setting monitoring condition to match real time data
Returned as monitoring result, for example, exception monitoring, daily record monitoring etc..Event monitoring also has identical monitoring logic, with monitoring
Exemplified by target is accident, accident should at least based on a supposed premise be that the event that occurred in history is present
It is also possible to occurring again, therefore, the keyword of historical events is the necessary factor of monitoring objective event condition.
Specifically, above-mentioned event keyword can be the set for the keyword that can describe certain a kind of event.Event is crucial
Word can be a message box, such as { time, place, main body, event word etc. }.In a kind of optional embodiment, for "
This event of shake ", can be described by " earthquake " keyword, can also by { China, earthquake }, Japan,
Shake } etc. { " place ", " event word " } be described；And for example, reported for beverage industry negative event, can be by { cola, killing
Essence }, { laughable, carcinogenic } etc. { " main body ", " event word " } is described.Message box as event keyword describing mode more
It is applicable, because this method for expressing has a more specific semanteme, the semantic reference more clearly to event is more unique.
Above-mentioned affair character information can be used for the dimension of description event keyword.In a kind of optional embodiment,
In same monitoring task, different events can be described using identical dimension, and during description, different event
What is differed is specific dimensional information.For example, in a monitoring task, the monitoring is used as using { " place ", " event " }
The dimension of task, then { " place ", " event word " } be used as event keyword, then { China, earthquake }, { Japan, earthquake }, { river in Shangdong Province
River, earthquake } it is exactly correct description, but { May 12, earthquake in 2008 } does not just describe for correct event keyword.And for example,
It is negative in above-mentioned beverage industry using the dimension of { " main body ", " event word " } as the monitoring task in a monitoring task
In the example of event report, event word " spermicidal ", " carcinogenic " are to add two features of negative emotion to be used as event using gerund part of speech
What characteristic information combined, therefore, { laughable, cause obesity } is exactly an event keyword, but { laughable, advertisement } or { can
It is happy, sad } just do not described for correct event keyword.
Herein it should be noted that being used for the history thing for obtaining event keyword and affair character information in above-mentioned steps
Part, it is the historical events corresponding with monitoring object, for example, in the case where monitoring object is accident, it is above-mentioned to be used to obtain
The event keyword and the historical events of affair character information taken is historical incident.
Step S104, the event keyword of historical events is trained, obtain the clause rule for monitoring online data.
In above-mentioned steps, clause rule can be the expression way of sentence.
The event either occurred in history, or new events, its general character is that they are all events, i.e., in event
It is roughly the same with being used for the reporting style of event although the description of event in itself is different on report.Therefore can be based on
The features described above of event, the expression way of event is modeled, with the description of extensive specific event keyword, obtains candidate's
Expression way.
Herein it should be noted that above-mentioned clause rule can be a variety of modes for being used to reporting or describing event, for example,
Clause rule can be in the clause of description event, the word before and after keyword, i.e., in the clause of description event, close
The position of keyword, using the word before and after keyword as a kind of clause rule, when also there is general description side in other clause
Formula, can the clause match with above-mentioned clause rule.
Step S106, according to clause rules detection online data, obtain candidate events.
, can be by matching clause expression way and text in online real time process in above-mentioned steps, can
To obtain the object of whole successful match.The clause that system will match to can be exported in the form of " event keyword ",
As candidate events.
Step S108, screened according to affair character information from candidate events and obtain object event.
It could be aware that, online data is monitored using clause rule, multiple things for meeting clause rule can be obtained
Part, but in the thing for meeting clause rule, it is also possible to including with other events of above-mentioned clause rule identical, therefore waited
After selecting event, the dimension of affair character information, i.e. keyword can be used, to filter out final mesh in multiple candidate events
Mark event.
Herein it should be noted that by artificial other means such as arrangement or log recording in advance, it can sum up and go through
The event keyword occurred in history, still, event (such as：Accident) not merely include what is had occurred and that in history
Event, in addition to some new events for not occurred in history.Therefore, it is difficult if be only monitored by event keyword
To monitor whole events.The scheme that the above embodiments of the present application provide is identical with the reporting style of historical events using new events
The characteristics of, the description of extensive specific event keyword, the clause rule obtained using the keyword of training historical events come to
Line number is according to being monitored, so as to ensure comprehensive, the accuracy of monitoring result.
From the foregoing, it will be observed that the application above-mentioned steps obtain the event keyword of historical events and affair character information, training are gone through
The event keyword of historical event part, the clause rule for monitoring online data is obtained, according to clause rules detection online data, is obtained
To candidate events, screened according to affair character information from candidate events and obtain object event.Such scheme can ensure to supervise
Survey promptness on the basis of, effectively improve it is above-mentioned for new accident None- identified the problem of, solve in the prior art, lead to
Keyword that overmatching is pre-set monitors accident, causes the low technical problem of accident monitoring accuracy.
Optionally, according to the above embodiments of the present application, the event keyword of historical events is trained, obtains being used to monitor online
The clause rule of data, including：
Step S1041, the clause rule of multiple first sentences is extracted, forms candidate's clause regular collection, wherein, Duo Ge
One sentence is that the obtained sentence for including whole event keywords in any historical events is searched in history language material.
In a kind of optional embodiment, still using the event of acquisition as accident as an example, in any historical events
For " Wenchuan occur 8.0 grades of special violent earthquakes ", in the case that event keyword is " Wenchuan " and " 8.0 ", comprising " Wenchuan " and
The clause of " 8.0 " can be used as the first sentence, and the clause for only including " 8.0 " comprising " Wenchuan " or only can not be used as the first language
Sentence.
Step S1043, the event keyword of the second sentence is extracted, wherein, the second sentence is in history language material and any
The clause of candidate's clause rule match.
, can be according to the event keyword of the second sentence of any candidate's clause Rule Extraction, with clause in above-mentioned steps
Rule as an example, include two placeholders { A } and { B } in the clause rule, is being gone through for " { A } occur { B } level special violent earthquake "
Searched in historical event part and obtain the sentence to match with the clause rule, wherein, in the sentence to match with the clause rule, with
Placeholder { A } and the event keyword that { B } corresponding word is the second sentence.
In a kind of optional embodiment, still using the event of acquisition as accident as an example, for clause rule
" { B } level special violent earthquake occurs for { A } ", " 8.0 grades of special violent earthquakes occur for Wenchuan " can be matched, " Japan occurs 9.0 grades especially bigly
The clause such as shake ", extract keyword { Wenchuan, earthquake }, { Japan, the earthquake } of the sentence of above-mentioned two second；For clause rule
" { A } { B } occur shooting incident ", it can match " April 26 shooting incident occurs for Texas ", extraction above-mentioned the
The keyword of two sentences is respectively and { Texas, April 26 days }.
Step S1045, in the event keyword for obtaining the second sentence, the first event keyword and second event keyword
Ratio, wherein, the first event keyword is the event keyword occurred in any historical events, and second event keyword is newly to go out
Existing keyword.
In a kind of optional embodiment, still using the event of acquisition as accident as an example, in this example, it is assumed that
" earthquake ", " Wenchuan " and " Texas " is the keyword occurred in any historical incident, and " April 26 " is to go through
History accident emerging keyword not to be covered, then in above three first, the first event keyword and the second thing
The ratio of part keyword is respectively 2:0,2:0 and 1:1.
Step S1047, in the case where the ratio of the first event keyword and second event keyword exceedes predetermined threshold value,
It is the clause rule for monitoring online data to confirm candidate's clause rule.
In a kind of optional embodiment, still using the event of acquisition as accident as an example it is supposed that in this example,
Above-mentioned predetermined threshold value is 1:1, then the first event keyword and second event in clause rule " { A } occur { B } level special violent earthquake "
The ratio of keyword is more than 1：1, the first event keyword and the second thing in clause rule " at { B } shooting incident occurs for { A } "
The ratio of part keyword is not less than 1:1, therefore, clause rule " at { B } shooting incident occurs for { A } " is eliminated, confirms that " { A } occurs
{ B } level special violent earthquake " is regular for the clause of monitoring online data.
Herein it should be noted that above-mentioned steps include whole events of any historical events by extracting in history language material
The clause rule of the sentence of keyword obtains multiple candidate's clause rules, then multiple candidate rules are screened, and obtaining can
For carrying out the clause rule of event monitoring, it is achieved thereby that the technical purpose of the event keyword in training historical events, and
Reach by training the keyword in historical events to obtain the clause rule for monitoring online data.
Herein it should also be noted that, above-mentioned steps using the ratio of the first event keyword and second event keyword as
Each candidate's clause rule is to the quantitative target of event description ability, if aforementioned proportion is larger, clause rule is to event
Descriptive power is stronger, on the contrary then weaker, therefore the time weaker to event description ability is eliminated by way of setting predetermined threshold value
Select clause regular, retain the candidate clause rule stronger to event description ability as the clause eventually for monitoring online data
Rule.
From the foregoing, it will be observed that the application above-mentioned steps obtain multiple candidate's clause rules, candidate's clause regular collection, extraction are formed
The event keyword of second sentence, wherein, the second sentence is the clause in history language material with any candidate's clause rule match,
In the event keyword for obtaining the second sentence, the ratio of the first event keyword and second event keyword, closed in the first event
In the case that the ratio of keyword and second event keyword exceedes predetermined threshold value, it is online for monitoring to confirm candidate's clause rule
The clause rule of data.Such scheme is fixed by way of calculating the ratio of the first event keyword and second event keyword
Amount judges descriptive power of candidate's clause rule to event, so as to be filtered out from multiple candidate's clause rules to event description energy
The stronger clause rule of power, as the clause rule eventually for monitoring online data.
Optionally, according to the above embodiments of the present application, multiple candidate's clause rules is obtained, form candidate's clause rule set
Close, including：
Step S1049, the first sentence is searched in history language material, wherein, the first sentence includes the complete of any historical events
Portion's event keyword.
In a kind of optional embodiment, still using the event of acquisition as accident as an example, in any historical events
For " Wenchuan occur 8.0 grades of special violent earthquakes ", in the case that event keyword is " Wenchuan " and " 8.0 ", comprising " Wenchuan " and
The clause of " 8.0 " can be used as the first sentence, and the clause for only including " 8.0 " comprising " Wenchuan " or only can not be used as the first language
Sentence.
Step S1051, the clause rule of multiple first sentences is extracted, forms candidate's clause regular collection.
From the foregoing, it will be observed that the application above-mentioned steps search the first sentence in history language material, the sentence of multiple first sentences is extracted
Formula rule, forms candidate's clause regular collection, wherein, the first sentence includes whole event keywords of any historical events.On
Scheme is stated by searching the first sentence in history language material, obtains candidate's clause rule.
Optionally, according to the above embodiments of the present application, according to clause rules detection online data, candidate events are obtained, are wrapped
Include：
Step S1061, the text in online data is matched with clause rule.
Step S1063, confirm that the text that the match is successful is candidate events.
In a kind of optional embodiment, still using the event of acquisition as accident as an example, in multiple clause rules
A clause rule for { A } occur { B } level special violent earthquake, matching when can with successful match to for example " Wenchuan generation 8.0 grades
Special violent earthquake (event keyword is { Wenchuan, earthquake }) ", " Japan occur 9.0 grades of special violent earthquakes (event keyword for Japan,
Earthquake }) ", candidate's accident such as " 8.0 grades of special violent earthquakes (event keyword is { yesterday, earthquake }) occur for yesterday ".
From the foregoing, it will be observed that the application above-mentioned steps are matched the text in online data with clause rule, confirm
The text that the match is successful is candidate events.Such scheme is using clause rule to the text progress in online data
Match somebody with somebody, so as to obtain candidate events.
Optionally, according to the above embodiments of the present application, the text in online data is matched with clause rule,
Including：
Step S1065, obtain text and each successful matching probability of clause rule match.
Step S1067, according to the order of matching probability from large to small, text is matched with clause rule.
It understand that by above-mentioned steps, matching process often has more than 1 clause rule, and text matches
Any one clause rule can be regarded as monitoring event, and therefore, matching process can be optimized.
In a kind of optional embodiment, still using the event of acquisition as accident as an example, for example there is 3 clause rule
Then A, B, C, pre-set text file then confirm as candidate events with any one clause rule successful match text, return to TRUE,
Otherwise it is assumed that it fails to match, FALSE is returned.Preset with text and A, the probability of B, C matching is respectively 20%, 50% He
80%, if being matched with A, B, C of acquiescence order, under aspiration level, need the number matched to be altogether
0.2*1+0.5* (1-0.2) * 2+1* (1-0.2-0.5* (1-0.2)) 3=2.2 times/bars of *.
Above-mentioned formula represents that, first by clause rule A, only 20% probability text is only once with regard to successful match
And T is returned, other 80% texts are required for continuing to match with clause rule B or C, when being matched with clause rule B, remain
Remaining in 80% text and have 50% (i.e. the 40% of sum) successful match and return to T, this 40% text have matched 2 altogether
It is secondary to obtain result, and also remaining 40% text then needs matching the 3rd time, and finally determine to return to T or F.
If the order of matching be optimized, i.e. C, B, A order is updated to, then the number of Mean match will be reduced
To 1.3 times/bar,
(0.8*1+0.5* (1-0.8) * 2+1* (1-0.8-0.5* (1-0.8)) * 3)=1.3 times/bar.Therefore, can be with sentence
Formula rule match can efficiently reduce system totality to the sequencing of how much decision clause matchings of accident probability
Matching times, significantly improve systematic function.
From the foregoing, it will be observed that the application above-mentioned steps obtain text and each successful matching probability of clause rule match,
According to the order of matching probability from large to small, text is matched with clause rule.Such scheme is by setting event
With the order of clause rule match, so as to reduce the number of Mean match, further ensureing the accuracy of acquisition event
In the case of, improve the ageing of acquisition event.
Optionally, according to the above embodiments of the present application, after candidate events are obtained, exported and waited in the form of event keyword
Select event.
In a kind of optional embodiment, still using the event of acquisition as accident as an example, presetting multiple candidate's things
Candidate events in part are " 8.0 grades of special violent earthquakes occur for Wenchuan ", or " 9.0 grades of special violent earthquakes occur for Japan ", respectively with
Event key word form exports to obtain " Wenchuan, earthquake ", and " Japan, earthquake ".
From the foregoing, it will be observed that the application above-mentioned steps export the candidate events got in the form of event keyword, so as to
Allow to match the event keyword of candidate events using affair character information, and then from multiple candidate events
Obtain object event.
Optionally, according to the above embodiments of the present application, screened according to affair character information from candidate events and obtain target
Event, including：
Step S1081, judge candidate events affair character information whether with clause rule affair character information phase
Match somebody with somebody.
Step S1083, confirm the affair character information and the affair character information match of clause rule of candidate events
Candidate events are object event.
It could be aware that by step S1041 to step S1047, although clause rule can efficiently identify event, clause
Rule is extracted to obtain with certain experiences threshold value, and therefore, in processing text data online, clause rule may be fitted on
The information of event keyword is not belonging to, therefore, candidate events can not export directly as final monitoring result, but need again
Carry out a step whether be currently needed for obtain event (such as：Accident) inspection.
Due to no matter historical events or new events, should all meet the dimension of affair character information.Therefore, from candidate events
In find real accident, then become two classification problems characterized by affair character information, no matter use rule
Then or any one sorting technique can be handled effectively the classification problem, following with a kind of optional embodiment pair
A kind of optional method is described.
In a kind of optional embodiment, still by accident of the event of acquisition as an example, such as clause rule is
{ B } level special violent earthquake occurs for { A }, and its affair character information is { A, earthquake }, can be with successful match to " Wenchuan occurs in matching
8.0 grades of special violent earthquakes (event keyword is { Wenchuan, earthquake }) ", " 9.0 grades of special violent earthquakes occur for Japan, and (event keyword is
{ Japan, earthquake }) ", " 8.0 grades of special violent earthquakes (event keyword is { yesterday, earthquake }) occur for yesterday " three candidates' burst things
Part.When the affair character information of definition is { place, event word=" earthquake " }, because the type of " yesterday " is not " place ",
Therefore this information filtering is fallen, and { Wenchuan, earthquake } and { Japan, earthquake } will be returned successfully.Also, when follow-up text
Meet " { A } generation { B } level special violent earthquake=>{ A, earthquake } " and " Tag (A)=place ", then just can successfully return.
From the foregoing, it will be observed that the application above-mentioned steps judge candidate events affair character information whether with clause rule event
Characteristic information matches, and confirms affair character information and the candidate of the affair character information match of clause rule of candidate events
Event is object event.Above-mentioned steps are screened by using affair character information to candidate events, have been filtered out by clause
The candidate events that rule is screened by mistake, obtain finally needing the object event obtained, so as to further solve in the prior art, lead to
Keyword that overmatching is pre-set monitors accident, causes the low technical problem of accident monitoring accuracy.
Embodiment 2
According to embodiments of the present invention, a kind of monitoring device of event is additionally provided, Fig. 2 is according to embodiments of the present invention 2
A kind of schematic diagram of the monitoring device of event, as shown in Fig. 2 the device can include：
First acquisition module 20, for obtaining the event keyword and affair character information of historical events, wherein, event is special
Reference ceases the dimensional information for characterizing event keyword.
Training module 22, for training the event keyword of historical events, obtain the clause for monitoring online data and advise
Then.
Monitoring modular 24, for according to clause rules detection online data, obtaining candidate events.
Screening module 26, object event is obtained for being screened according to affair character information from candidate events.
Herein it should be noted that by artificial other means such as arrangement or log recording in advance, it can sum up and go through
The event keyword occurred in history, still, event (such as：Accident) not merely include what is had occurred and that in history
Event, in addition to some new events for not occurred in history.Therefore, it is difficult if be only monitored by event keyword
To monitor whole events.The scheme that the above embodiments of the present application provide is identical with the reporting style of historical events using new events
The characteristics of, the description of extensive specific event keyword, the clause rule obtained using the keyword of training historical events come to
Line number is according to being monitored, so as to ensure comprehensive, the accuracy of monitoring result.
From the foregoing, it will be observed that the application said apparatus obtains the event keyword and event of historical events by the first acquisition module
Characteristic information, the event keyword of historical events is trained by training module, obtains the clause rule for monitoring online data,
By monitoring modular according to clause rules detection online data, candidate events are obtained, are believed by screening module according to affair character
Breath screening from candidate events obtains object event.Such scheme can effectively improve on the basis of ensureing to monitor promptness
It is above-mentioned for new accident None- identified the problem of, solve in the prior art, by match the keyword that pre-sets come
Accident is monitored, causes the low technical problem of accident monitoring accuracy.
Optionally, according to the above embodiments of the present application, with reference to shown in Fig. 3, above-mentioned training module 22 includes：
Abstraction module 30, for extracting the clause rule of multiple first sentences, candidate's clause regular collection is formed, wherein,
Multiple first sentences are that the obtained sentence for including whole event keywords in any historical events is searched in history language material.
Extraction module 32, for extract the second sentence event keyword, wherein, the second sentence be in history language material with
The clause of any candidate's clause rule match.
Second acquisition module 34, for obtain the second sentence event keyword in, the first event keyword and the second thing
The ratio of part keyword, wherein, the first event keyword is the event keyword occurred in any historical events, and second event closes
Keyword is emerging keyword.
First confirms module 36, exceedes predetermined threshold value for the ratio in the first event keyword and second event keyword
In the case of, it is the clause rule for monitoring online data to confirm candidate's clause rule.
From the foregoing, it will be observed that the application said apparatus obtains multiple candidate's clause rules by the second acquisition module, candidate is formed
Clause regular collection, the event keyword of the second sentence is extracted by extraction module, wherein, the second sentence is in history language material
With the clause of any candidate's clause rule match, in the event keyword that the second sentence is obtained by the 3rd acquisition module, first
The ratio of event keyword and second event keyword, exceed in the ratio of the first event keyword and second event keyword pre-
If in the case of threshold value, confirm that candidate's clause rule is the clause rule for monitoring online data by the first confirmation module.
Such scheme is by way of calculating the ratio of the first event keyword and second event keyword, rational judgment candidate clause rule
Then to the descriptive power of event, so as to filter out the clause rule stronger to event description ability from multiple candidate's clause rules
Then, as the clause rule eventually for monitoring online data.
Optionally, according to the above embodiments of the present application, with reference to shown in Fig. 4, monitoring modular 24 includes：
Matching module 40, for the text in online data to be matched with clause rule；
Second confirms module 42, for confirming that the text that the match is successful is candidate events.
From the foregoing, it will be observed that the application said apparatus is entered the text in online data and clause rule by matching module
Row matching, confirm that the text that the match is successful is candidate events by matching module.Such scheme using clause rule to
Text of the line number in is matched, so as to obtain candidate events.
Optionally, according to the above embodiments of the present application, with reference to shown in Fig. 5, matching module 40 includes：
3rd acquisition module 50, for obtaining text and each successful matching probability of clause rule match；
Matched sub-block 52, for the order according to matching probability from large to small, text and clause rule are carried out
Matching.
From the foregoing, it will be observed that the application said apparatus obtains text and each clause rule match by the 4th acquisition module
Successful matching probability, by order of the matched sub-block according to matching probability from large to small, by text and clause rule
Matched.Such scheme, so as to reduce the number of Mean match, enters one by setting the order of event and clause rule match
Step ensure acquisition event accuracy in the case of, improve the ageing of acquisition event.
Optionally, according to the above embodiments of the present application, after candidate events are obtained, exported and waited in the form of event keyword
Select event.
From the foregoing, it will be observed that the application said apparatus exports the candidate events got in the form of event keyword, so as to
Allow to match the event keyword of candidate events using affair character information, and then from multiple candidate events
Obtain object event.
Optionally, according to the above embodiments of the present application, with reference to shown in Fig. 6, screening module 26 includes：
Judge module 60, for judge candidate events affair character information whether with clause rule affair character information
Match；
3rd confirms module 62, the affair character information of affair character information and clause rule for confirming candidate events
The candidate events to match are object event.
From the foregoing, it will be observed that the application said apparatus by judge module judge candidate events affair character information whether with sentence
The affair character information match of formula rule, confirm that module confirms that the affair character information of candidate events and clause are advised by the 3rd
The candidate events of affair character information match then are object event.Such scheme is by using affair character information to candidate
Event is screened, and has been filtered out the candidate events screened by mistake by clause rule, has been obtained finally needing the object event obtained, from
And further solve in the prior art, accident is monitored by matching the keyword pre-set, causes accident
Monitor the low technical problem of accuracy.
The monitoring device of event includes processor and memory, above-mentioned first acquisition module, training module, monitoring modular and
Screening module etc. stores in memory as program unit, by the said procedure list of computing device storage in memory
Member realizes corresponding function.
Kernel is included in processor, is gone in memory to transfer corresponding program unit by kernel.Kernel can set one
Or more, solve in the prior art, burst thing to be monitored by matching the keyword pre-set by adjusting kernel parameter
Part, cause the low technical problem of accident monitoring accuracy.
Memory may include computer-readable medium in volatile memory, random access memory (RAM) and/
Or the form such as Nonvolatile memory, such as read-only storage (ROM) or flash memory (flash RAM), memory includes at least one deposit
Store up chip.
Present invention also provides a kind of computer program product, when being performed on data processing equipment, is adapted for carrying out just
The program code of beginningization there are as below methods step：The event keyword and affair character information of historical events are obtained, wherein, event
Characteristic information is used for the dimensional information for characterizing event keyword；The event keyword of historical events is trained, obtains being used to monitor
The clause rule of line number evidence；According to clause rules detection online data, candidate events are obtained；According to affair character information from candidate
Screening obtains object event in event.
The embodiments of the present invention are for illustration only, do not represent the quality of embodiment.
In the above embodiment of the present invention, the description to each embodiment all emphasizes particularly on different fields, and does not have in some embodiment
The part of detailed description, it may refer to the associated description of other embodiment.
In several embodiments provided herein, it should be understood that disclosed technology contents, others can be passed through
Mode is realized.Wherein, device embodiment described above is only schematical, such as the division of the unit, Ke Yiwei
A kind of division of logic function, can there is an other dividing mode when actually realizing, for example, multiple units or component can combine or
Person is desirably integrated into another system, or some features can be ignored, or does not perform.Another, shown or discussed is mutual
Between coupling or direct-coupling or communication connection can be INDIRECT COUPLING or communication link by some interfaces, unit or module
Connect, can be electrical or other forms.
The unit illustrated as separating component can be or may not be physically separate, show as unit
The part shown can be or may not be physical location, you can with positioned at a place, or can also be distributed to multiple
On unit.Some or all of unit therein can be selected to realize the purpose of this embodiment scheme according to the actual needs.
In addition, each functional unit in each embodiment of the present invention can be integrated in a processing unit, can also
That unit is individually physically present, can also two or more units it is integrated in a unit.Above-mentioned integrated list
Member can both be realized in the form of hardware, can also be realized in the form of SFU software functional unit.
If the integrated unit is realized in the form of SFU software functional unit and is used as independent production marketing or use
When, it can be stored in a computer read/write memory medium.Based on such understanding, technical scheme is substantially
The part to be contributed in other words to prior art or all or part of the technical scheme can be in the form of software products
Embody, the computer software product is stored in a storage medium, including some instructions are causing a computer
Equipment (can be personal computer, server or network equipment etc.) perform each embodiment methods described of the present invention whole or
Part steps.And foregoing storage medium includes：USB flash disk, read-only storage (ROM, Read-Only Memory), arbitrary access are deposited
Reservoir (RAM, Random Access Memory), mobile hard disk, magnetic disc or CD etc. are various can be with store program codes
Medium.
Described above is only the preferred embodiment of the present invention, it is noted that for the ordinary skill people of the art
For member, under the premise without departing from the principles of the invention, some improvements and modifications can also be made, these improvements and modifications also should
It is considered as protection scope of the present invention.
Claims (10)
- A kind of 1. monitoring method of event, it is characterised in that including：The event keyword and affair character information of historical events are obtained, wherein, the affair character information is described for characterizing The dimensional information of event keyword；The event keyword of the historical events is trained, obtains the clause rule for monitoring online data；According to online data described in the clause rules detection, candidate events are obtained；Screened according to the affair character information from the candidate events and obtain object event.
- 2. according to the method for claim 1, it is characterised in that train the event keyword of the historical events, used In the clause rule of monitoring online data, including：Extract multiple first sentences clause rule, form candidate's clause regular collection, wherein, the multiple first sentence be The obtained sentence for including whole event keywords in any historical events is searched in history language material；The event keyword of the second sentence is extracted, wherein, second sentence is in the history language material and any time Select the sentence of clause rule match；In the event keyword for obtaining second sentence, the ratio of the first event keyword and second event keyword, wherein, The first event keyword is the event keyword occurred in any historical events, and the second event keyword occurs to be new Keyword；In the case where the ratio of the first event keyword and the second event keyword exceedes predetermined threshold value, institute is confirmed It is the clause rule for monitoring online data to state candidate's clause rule.
- 3. according to the method for claim 1, it is characterised in that according to online data described in the clause rules detection, obtain To candidate events, including：Text in the online data is matched with the clause rule；Confirm that the text that the match is successful is the candidate events.
- 4. according to the method for claim 3, it is characterised in that by the text in the online data and the clause Rule is matched, including：Obtain the text and each successful matching probability of clause rule match；According to the order of the matching probability from large to small, the text is matched with the clause rule.
- 5. according to the method for claim 4, it is characterised in that according to the affair character information from the candidate events Screening obtains current event, including：Judge the candidate events affair character information whether with the clause rule affair character information match；Confirm the affair character information and the time of the affair character information match of clause rule of the candidate events It is the object event to select event.
- A kind of 6. monitoring device of event, it is characterised in that including：First acquisition module, for obtaining the event keyword and affair character information of historical events, wherein, the affair character Information is used for the dimensional information for characterizing the event keyword；Training module, for training the event keyword of the historical events, obtain the clause rule for monitoring online data；Monitoring modular, for the online data according to the clause rules detection, obtain candidate events；Screening module, object event is obtained for being screened according to the affair character information from the candidate events.
- 7. device according to claim 6, it is characterised in that the training module includes：Abstraction module, for extracting the clause rule of multiple first sentences, candidate's clause regular collection is formed, wherein, it is described more Individual first sentence is that the obtained sentence for including whole event keywords in any historical events is searched in history language material；Extraction module, for extracting the event keyword of the second sentence, wherein, second sentence is in the history language material With the clause of any candidate's clause rule match；Second acquisition module, in the event keyword for obtaining second sentence, the first event keyword and second event The ratio of keyword, wherein, the first event keyword is the event keyword that occurs in any historical events, described second Event keyword is emerging keyword；First confirms module, for exceeding default threshold in the ratio of the first event keyword and the second event keyword In the case of value, it is the clause rule for monitoring online data to confirm candidate's clause rule.
- 8. device according to claim 6, it is characterised in that the monitoring modular includes：Matching module, for the text in the online data to be matched with the clause rule；Second confirms module, for confirming that the text that the match is successful is the candidate events.
- 9. device according to claim 8, it is characterised in that the matching module includes：3rd acquisition module, for obtaining the text and each successful matching probability of clause rule match；Matched sub-block, for the order according to the matching probability from large to small, the text and the clause are advised Then matched.
- 10. device according to claim 9, it is characterised in that the screening module includes：Judge module, for judge the candidate events affair character information whether with the clause rule affair character believe Manner of breathing matches；3rd confirms module, for confirming that the affair character of affair character information and clause rule of the candidate events is believed The candidate events of manner of breathing matching are the object event.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610659660.XA CN107729337B (en) | 2016-08-11 | 2016-08-11 | Event monitoring method and device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610659660.XA CN107729337B (en) | 2016-08-11 | 2016-08-11 | Event monitoring method and device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN107729337A true CN107729337A (en) | 2018-02-23 |
| CN107729337B CN107729337B (en) | 2020-05-22 |
Family
ID=61199731
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610659660.XA Active CN107729337B (en) | 2016-08-11 | 2016-08-11 | Event monitoring method and device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN107729337B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109543185A (en) * | 2018-11-22 | 2019-03-29 | 联想(北京)有限公司 | Utterance topic acquisition methods and device |
| CN110162712A (en) * | 2019-05-28 | 2019-08-23 | 武汉轻工大学 | Event acquisition methods, device, equipment and storage medium |
| CN111339766A (en) * | 2020-02-19 | 2020-06-26 | 云南电网有限责任公司昆明供电局 | Operation ticket compliance checking method and device |
| CN111507110A (en) * | 2019-01-30 | 2020-08-07 | 国家计算机网络与信息安全管理中心 | Method, device and equipment for detecting emergency and storage medium |
| CN117669513A (en) * | 2024-01-30 | 2024-03-08 | 江苏古卓科技有限公司 | Data management system and method based on artificial intelligence |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070233650A1 (en) * | 2006-03-29 | 2007-10-04 | Chad Brower | Automatic categorization of network events |
| US20090024469A1 (en) * | 2007-07-20 | 2009-01-22 | Andrei Zary Broder | System and Method to Facilitate Classification and Storage of Events in a Network |
| CN101937430A (en) * | 2010-09-03 | 2011-01-05 | 清华大学 | Method for extracting event sentence pattern from Chinese sentence |
| CN102610101A (en) * | 2012-04-01 | 2012-07-25 | 北京世纪高通科技有限公司 | Method for collecting information of traffic incidents |
| CN103150432A (en) * | 2013-03-07 | 2013-06-12 | 宁波成电泰克电子信息技术发展有限公司 | Method for internet public opinion analysis |
| CN104091054A (en) * | 2014-06-26 | 2014-10-08 | 中国科学院自动化研究所 | Mass disturbance warning method and system applied to short texts |
| CN104252488A (en) * | 2013-06-28 | 2014-12-31 | 华为技术有限公司 | Data processing method and server |
-
2016
- 2016-08-11 CN CN201610659660.XA patent/CN107729337B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070233650A1 (en) * | 2006-03-29 | 2007-10-04 | Chad Brower | Automatic categorization of network events |
| US20090024469A1 (en) * | 2007-07-20 | 2009-01-22 | Andrei Zary Broder | System and Method to Facilitate Classification and Storage of Events in a Network |
| CN101937430A (en) * | 2010-09-03 | 2011-01-05 | 清华大学 | Method for extracting event sentence pattern from Chinese sentence |
| CN102610101A (en) * | 2012-04-01 | 2012-07-25 | 北京世纪高通科技有限公司 | Method for collecting information of traffic incidents |
| CN103150432A (en) * | 2013-03-07 | 2013-06-12 | 宁波成电泰克电子信息技术发展有限公司 | Method for internet public opinion analysis |
| CN104252488A (en) * | 2013-06-28 | 2014-12-31 | 华为技术有限公司 | Data processing method and server |
| CN104091054A (en) * | 2014-06-26 | 2014-10-08 | 中国科学院自动化研究所 | Mass disturbance warning method and system applied to short texts |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109543185A (en) * | 2018-11-22 | 2019-03-29 | 联想(北京)有限公司 | Utterance topic acquisition methods and device |
| CN109543185B (en) * | 2018-11-22 | 2021-11-16 | 联想(北京)有限公司 | Statement topic acquisition method and device |
| CN111507110A (en) * | 2019-01-30 | 2020-08-07 | 国家计算机网络与信息安全管理中心 | Method, device and equipment for detecting emergency and storage medium |
| CN110162712A (en) * | 2019-05-28 | 2019-08-23 | 武汉轻工大学 | Event acquisition methods, device, equipment and storage medium |
| CN110162712B (en) * | 2019-05-28 | 2022-06-07 | 武汉轻工大学 | Event acquisition method, device, equipment and storage medium |
| CN111339766A (en) * | 2020-02-19 | 2020-06-26 | 云南电网有限责任公司昆明供电局 | Operation ticket compliance checking method and device |
| CN117669513A (en) * | 2024-01-30 | 2024-03-08 | 江苏古卓科技有限公司 | Data management system and method based on artificial intelligence |
| CN117669513B (en) * | 2024-01-30 | 2024-04-12 | 江苏古卓科技有限公司 | Data management system and method based on artificial intelligence |
Also Published As
| Publication number | Publication date |
|---|---|
| CN107729337B (en) | 2020-05-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107729337A (en) | The monitoring method and device of event | |
| CN107861942B (en) | Suspected power complaint work order identification method based on deep learning | |
| JP7090936B2 (en) | ESG-based corporate evaluation execution device and its operation method | |
| CN104598367B (en) | Data center's event of failure management specialty and method | |
| CN109543032A (en) | File classification method, device, computer equipment and storage medium | |
| CN110532451A (en) | Search method and device for policy text, storage medium, electronic device | |
| CN107437038A (en) | A kind of detection method and device of webpage tamper | |
| CN109614550A (en) | Public sentiment monitoring method, device, computer equipment and storage medium | |
| CN110377731A (en) | Complain text handling method, device, computer equipment and storage medium | |
| CN110134845A (en) | Project public sentiment monitoring method, device, computer equipment and storage medium | |
| CN111177382B (en) | Intelligent legal system recommendation auxiliary system based on FastText algorithm | |
| CN108491388A (en) | Data set acquisition methods, sorting technique, device, equipment and storage medium | |
| CN108229772A (en) | Mark processing method and processing device | |
| CN106649334A (en) | Conjunction word set processing method and device | |
| CN110138794A (en) | A kind of counterfeit website identification method, device, equipment and readable storage medium storing program for executing | |
| CN112116168B (en) | User behavior prediction method and device and electronic equipment | |
| CN110134844A (en) | Subdivision field public sentiment monitoring method, device, computer equipment and storage medium | |
| KR102650139B1 (en) | Artificial intelligence-based system and method for online counterfeit product crackdown | |
| CN110458296A (en) | The labeling method and device of object event, storage medium and electronic device | |
| CN108984514A (en) | Acquisition methods and device, storage medium, the processor of word | |
| CN113407644A (en) | Enterprise industry secondary industry multi-label classifier based on deep learning algorithm | |
| CN116865994A (en) | Network data security prediction method based on big data | |
| CN112347254A (en) | News text classification method and device, computer equipment and storage medium | |
| CN106846170B (en) | Generator set trip monitoring method and monitoring device thereof | |
| CN109446299A (en) | The method and system of searching email content based on event recognition |
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 | ||
| CB02 | Change of applicant information | ||
| CB02 | Change of applicant information |
Address after: 100083 No. 401, 4th Floor, Haitai Building, 229 North Fourth Ring Road, Haidian District, Beijing Applicant after: Beijing Guoshuang Technology Co.,Ltd. Address before: 100086 Cuigong Hotel, 76 Zhichun Road, Shuangyushu District, Haidian District, Beijing Applicant before: Beijing Guoshuang Technology Co.,Ltd. |
|
| GR01 | Patent grant | ||
| GR01 | Patent grant |