Patent Grant
8296045
1. Field of Invention
The present invention relates to the technical field of traffic information processing, and in particular to a traffic information processing apparatus and method as well as a traffic information integrating device and method.
2. Description of Prior Art
Nowadays, traffic congestion has increasingly become one of the major issues in modern cities due to the rapid advance of current auto industry and people's reliance on such vehicles. So far, efforts have been made in finding an effective solution for traffic jams. The existing techniques attempt to alleviate traffic congestions by providing traffic information service, in which the traffic information can be gathered from a plurality of heterogeneous information sources, such as traffic floating car, sensor/camera, navigating system, user report and the like. Although a great deal of traffic information can be gathered from multiple heterogeneous information sources, it is difficult to make full use of the traffic information since the information originates from various sources, and thus there are usually many kinds of differences among the gathered information, such as format variation, synonymous problem, spatial inconsistence, temporal inconsistence and semantic inconsistence. Therefore, how to efficiently integrate such heterogeneous traffic information has come to be an important and imperative issue.
Some systems and methods have been proposed for integrating traffic information from different information sources.
JP2006023886 reveals a traffic information system which acquires traffic information from multiple information sources as well as decimates and merges the acquired traffic information according to different route sections.
U.S. Pat. No. 5,173,691 discloses a method for fusing traffic congestion information, which can process raw traffic information from various sources and transform the raw information into traffic descriptions for respective route sections. If multiple traffic descriptions exist with respect to the same route section, selection of traffic information is conducted in accordance with the reliability of respective information sources.
JP2002260162 provides a traffic information provision system which can provide comprehensive global traffic information based on local traffic information transmitted from multiple mobile phone terminals. If several traffic descriptions exist with respect to the same district, selection of traffic information is conducted in accordance with the reliability of respective information sources.
In summary, the known traffic information processing methods or provision systems can perform simple combination of several heterogeneous traffic information as well as process traffic information having different formats or being synonymous with each other. Unfortunately, for the gathered traffic information, these known methods and systems cannot process the part that has inconsistence in terms of space, time or semantics, and thus cannot provide users with accurate, complete and reliable traffic information.
The present invention is made to solve the above problems. The present invention provides a traffic information processing apparatus and method, by which traffic information having different formats, the same meaning or inconsistence in space, time or semantics can be processed so as to integrate effectively the traffic information from a variety of heterogeneous information sources and ensure the accuracy, completeness and reliability of traffic data.
According to the first aspect of the present invention, a traffic information processing apparatus is provided comprising:
a format unifying device which unifies input traffic information with different formats to traffic information with unified format; and
a traffic information integrating device which corrects and/or complements the traffic information with unified format based on a knowledge base to obtain traffic information which is consistent with each other, so as to integrate the traffic information, wherein the knowledge base is external to the apparatus or internal to the apparatus.
According to the second aspect of the present invention, the traffic information integrating device comprises:
an input unit which receives the traffic information with unified format;
a spatial conflict processing unit which detects the traffic information of which described spatial locations are adjoined but the traffic description are contradict with each other according to a relation section of the knowledge base, and corrects the contradict traffic information so as to integrate the traffic information; and
an output unit which outputs the integrated traffic information.
According to the third aspect of the present invention, the traffic information integrating device comprises:
an input unit which receives the traffic information with unified format;
a spatial complementary processing unit which detects the traffic information of which described spatial location are adjoined but the traffic description are not complete according to a relation section of the knowledge base, and then generates complementary traffic information so as to form the integrated traffic information;
an output unit which outputs the integrated traffic information.
According to the fourth aspect of the present invention, the traffic information integrating device comprises:
an input unit which receives the traffic information with unified format;
a time conflict processing unit which corrects the traffic information that has time conflict with each other by comparing the traffic status about the same location on different time according to the concept section in the knowledge base, so as to integrate the traffic information; and
an output unit which outputs the integrated traffic information.
According to the fifth aspect of the present invention, the traffic information integrating device comprises:
an input unit which receives the traffic information with unified format;
a time complementary processing unit which adds new traffic information or complements the traffic information that lacks traffic status by comparing the traffic information on adjacent time, so as to integrate the traffic information; and
an output unit which outputs the integrated traffic information.
According to the sixth aspect of the present invention, the traffic information integrating device comprises:
an input unit which receives the traffic information with unified format;
a semantic conflict processing unit which searches the traffic information that has semantic conflict with each other based on the relation section in the knowledge base, and selects the traffic information having high reliability from the traffic information that has semantic conflict with each other, according to at least one of the determination conditions including the reliability of an information source, the majority having priority, comparison with the current time and comparison with history traffic data, so as to integrate the traffic information; and
an output unit which outputs the integrated traffic information.
According to the seventh aspect of the present invention, the traffic information integrating device comprises:
an input unit which receives the traffic information with unified format;
a semantic complementary processing unit which searches the traffic information that semantically complements with each other according to the relation section in the knowledge base, and combines the traffic information that semantically complements with each other, so as to integrate the traffic information; and
an output unit which outputs the integrated traffic information.
According to the eighth aspect of the present invention, a traffic information processing method is provided comprising:
a format unifying step of unifying input traffic information with different formats to traffic information with unified format; and
a traffic information integrating step of correcting and/or complementing the traffic information with unified format based on a knowledge base to obtain traffic information which is consistent with each other, so as to integrate the traffic information.
According to the ninth aspect of the present invention, the traffic information integrating steps comprises at least one of the following steps:
a spatial conflict processing step of detecting the traffic information of which described spatial location are adjoined but the traffic description are contradict with each other according to a relation section of the knowledge base, and correcting the contradict traffic information so as to integrate the traffic information;
a spatial complementary processing step of detecting the traffic information of which described spatial location are adjoined but the traffic description are not complete according to a relation section of the knowledge base, and then generating complementary traffic information so as to form the integrated traffic information;
a time conflict processing step of correcting the traffic information that has time conflict with each other by comparing the traffic status about the same location on different time according to the concept section in the knowledge base, so as to integrate the traffic information; and;
a time complementary processing step of adding new traffic information or complementing the traffic information that lacks traffic status by comparing the traffic information on adjacent time, so as to integrate the traffic information;
a semantic conflict processing step of searching the traffic information that has semantic conflict with each other based on the relation section in the knowledge base, and selecting the traffic information having high reliability from the traffic information that has semantic conflict with each other, according to at least one of the determination conditions including the reliability of an information source, the majority having priority, comparison with the current time and comparison with history traffic data, so as to integrate the traffic information;
a semantic complementary processing step of searching the traffic information that semantically complements with each other according to the relation section in the knowledge base, combining the traffic information that semantically complements with each other, so as to integrate the traffic information.
According to the tenth aspect of the present invention, a traffic information processing apparatus is provided comprising:
an input device, which receives traffic information with unified format;
a traffic information integrating device which corrects and/or complements the traffic information with unified format based on knowledge base to obtain traffic information which is consistent with each other, so as to integrate the traffic information, wherein the knowledge base is external to the apparatus or internal to the apparatus; and
an output device which outputs the integrated traffic information.
According to the eleventh aspect of the present invention, a traffic information processing method is provided comprising:
an input step of receiving traffic information with unified format; and
a traffic information integrating step of correcting and/or complementing the traffic information with unified format based on knowledge base to obtain traffic information which is consistent with each other, so as to integrate the traffic information; and
an output step of outputting the integrated traffic information.
According to the twelfth aspect of the present invention, a traffic information processing method is provided, comprising at least one of the following steps:
a spatial conflict processing step of detecting the traffic information of which described spatial location are adjoined but the traffic description are contradict with each other according to a relation section of the knowledge base, and correcting the contradict traffic information so as to integrate the traffic information;
a spatial complementary processing step of detecting the traffic information of which described spatial location are adjoined but the traffic description are not complete according to a relation section of the knowledge base, and then generating complementary traffic information so as to form the integrated traffic information;
a time conflict processing step of correcting the traffic information that has time conflict with each other by comparing the traffic status about the same location on different time according to the concept section in the knowledge base, so as to integrate the traffic information; and;
a time complementary processing step of adding new traffic information or complementing the traffic information that lacks traffic status by comparing the traffic information on adjacent time, so as to integrate the traffic information;
a semantic conflict processing step of searching the traffic information that has semantic conflict with each other based on the relation section in the knowledge base, and selecting the traffic information having high reliability from the traffic information that has semantic conflict with each other, according to at least one determination condition from the determination conditions including the reliability of an information source, the majority having priority, comparison with the current time and comparison with history traffic data, so as to integrate the traffic information; and
a semantic complementary processing step of searching the traffic information that semantically complements with each other according to the relation section in the knowledge base, and combining the traffic information that semantically complements with each other, so as to integrate the traffic information.
a) shows an exemplary concept table of the knowledge base according to the present invention;
b) shows an exemplary relation table of the knowledge base according to the present invention;
a) shows an example of traffic information spatial conflict processing according to the present invention;
b) shows an example of traffic information spatial complement processing according to the present invention;
a) shows an example of traffic information time conflict processing according to the present invention;
b) shows an example of traffic information time complement processing according to the present invention;
Now, a description will be give to the preferred embodiments of the present invention with reference to the figures, throughout which the same elements are denoted by the same reference symbols or numbers. Besides, in the following description, detailed explanation of known functions and configurations will not be repeated, otherwise it may obscure the subject of the present invention.
Particularly, the traffic information is received and transmitted by the input device 101 to the format unifying device 102, which unifies the formats of the traffic information and then transmits it to the synonymous information processing device 103. The synonymous information processing device 103 transforms the synonymous descriptions in the traffic information with unified format to identical description information and transmits the transformed traffic information to the checking device 104, which checks the received traffic information and transmits the checked traffic information to the knowledge base mapping device 105. The traffic information is received and mapped into the traffic information defined in the knowledge base 108 by the knowledge base mapping device 105 and transmitted to the traffic information integrating device 106. The traffic information integrating device 106 integrates the traffic information with spatial, temporal and semantic inconsistence according to the knowledge base 108. After that, the output device 107 is adapted to output the integrated traffic information.
It will be appreciated that the above illustrates only an exemplary traffic information processing apparatus. Alternatively, each of the format unifying device 102, the synonymous information processing device 103, the check device 104 and the knowledge base mapping device 105 can be optional device in the present invention.
The traffic information processing apparatus of the present invention analyzes and processes the received traffic information according to the knowledge base 108 so as to provide accurate and complete traffic information with spatial, temporal and semantic consistence. Thus, the knowledge base 108 according to the present invention will be first introduced with reference to
As shown in
(1) The concept section defines various concepts associated with traffic situations. The types of concepts include geographic categories, geographic entities, geographic direction and traffic status. The geographic entities usually have three types: point, line and plane, in which the point contains a bridge, an intersection and a point of interest (geographic object), such as the Xueyuan Bridge, the Hailong Building and the Beihang West Gate; the line consists of a road, such as the Zhichun Road and the Fourth North Ring Road; and the plane comprises a region, such as Tsinghua University and the Zhongguancun District. Each concept is represented as (name, type, synonymous), and the type and all the synonyms are specified for each concept. All concepts and their representations can be formed into a concept table.
(2) The attribute section defines the characteristics of a concept. For example, longitude and latitude are employed to depict the coordinates of a geographic entity in a map. Each attribute has at least one item “type” indicating the value type of the attribute. Any geographic entity can be precisely located with help of the attribute section.
(3) The relation section depicts various associations between concepts, that is, defines spatial relationship and semantic relationship between concepts. Each relation can be represented as (relation name, relation type, relation value type, relation value set, note), in which the relation name denotes the name of the relation, the relation type indicates whether the relation belongs to spatial relationship or a semantic relationship, the relation value type defines the value type of parameters of the relation, the relation value set represents all specific instances regarding the relation, and the note provides relevant explanation for the relation, i.e., explaining that certain relation represents a sequential relation, a subordinate relation or a causal relation. The spatial relationship comprises the sequential relation and location adjoining between geographic entities, and the semantic relationship contains such relations as subordinate, causal and antonymous relations. The relation type, relation value type and relation value set are specified for each relation, and the note can be specified. All relations and their representations can be combined into a relation table.
b) shows an example of the relation table of the knowledge base. The word “cross” depicts the sequential relation between geographic entities, for example, “cross (the Fourth North Ring Road, the BaofuShi Bridge, the Xueyuan Bridge, the Jiangxiang Bridge, . . . ,)” denotes that the Fourth North Ring Road routes along the BaofuShi Bridge, the Xueyuan Bridge and the Jiangxiang Bridge sequentially. The word “between” describes the location adjoining relation between geographic entities, for example, “between (the Xuezhi Bridge, the Xueyuan Bridge, the Jimen Bridge)” indicates that the Xuezhi Bridge lies between the Xueyuan Bridge and the Jimen Bridge. The word “isa” denotes the subordinate relation between geographic entities and geographic categories, for example, “isa (the Fourth North Ring Road, Road)” implicates that the Fourth North Ring Road is a road. The word “causal” illustrates the causal relation between traffic statuses, for example, “causal(congested, traffic accident)” indicates that the traffic accident results in the congestion. The word “antonym” means the antonymous relation between traffic statues, for example, “antonym(congested, unblocked)” points out that congested and unblocked are opposite to each other. Since each attribute or relation defines certain aspect of a concept, a plurality of attributes and corresponding relations can be combined to describe an integral concept.
(4) The axiom section is rules which are based on the concept section and the relation section and can further deduce the spatial relationship between concepts. For the axiom “cross (x, y1, . . . , yi, . . . )→between (yi, yi−1,yi+1)”, if the Fourth North Ring Road routes along the BaofuShi Bridge, the Xueyuan Bridge and the Jiangxiang Bridge sequentially, it can be deduced that the Xueyuan Bridge is located between the BaofuShi Bridge and the Jiangxiang Bridge, that is, “cross (the Fourth North Ring Road, the BaofuShi Bridge, the Xueyuan Bridge, the Jiangxiang Bridge)→between (the Xueyuan Bridge, the BaofuShi Bridge, the Jiangxiang Bridge)”. In general, the rules in the axiom section are limited in terms of number and can be expanded as demand.
Next, the method for generating the knowledge base 108 shown in
Normally, all locations and corresponding geographic categories as well as their position data (e.g., longitude, latitude) are taken from the existing electronic map, and a set of GIS functions is utilized to conduct spatial calculation based on these data. The history traffic database stores all the traffic information received from respective heterogeneous information sources by the traffic information processing apparatus. Each piece of traffic information contains at least location, traffic and time, and it can also contain information source and direction. The semantic dictionary is a known one, such as Hownet.
Now referring to
The attribute section is extracted at S302. The data fields (e.g., longitude, latitude) associated with respective location description are first extracted from respective data tables in the electronic map database as an attribute. Subsequently, the value type corresponding to each attribute is obtained according to the type definition of each data field. The attribute and the value type serve together as the attribute section.
The relation section is extracted at S303. Geographic categories to which respective geographic entities belong are first extracted from the electronic map database so as to obtain the subordinate relation between the geographic entities and the geographic categories, with the subordinate relation belonging to the semantic relationship. Next, the spatial relationship between the geographic entities is calculated using the existing GIS functions. Then, other semantic relationship between respective concepts is acquired by utilizing known semantic dictionaries like Hownet. As an example, “antonym (congested, unblocked)” is extracted on the basis that the word “congested” and the word “unblocked” are opposite to each other. Eventually, a relation table is generated in accordance with the above extract result.
At S304, the extracted concept section, attribute section, relation section and manually-defined axiom section are combined to form the ultimate knowledge base 108.
The knowledge base comprises the concept, attribute, relation and axiom sections, and the spatial and semantic relationships between concepts are defined in the relation section. Therefore, the traffic information processing apparatus according to the present invention can utilize the knowledge base to analyze the spatial relationship or the semantic relationship represented by the traffic information to be processed and thus can process the traffic information inconsistent in space, time and semantic, thereby providing accurate and complete traffic information with a high reliability.
While
At S602, the format unifying device 102 unifies the information description formats of the input traffic information with different formats.
At S603, the traffic information with unified format is transmitted to the synonymous information processing device 103, which searches the concept table in the knowledge base. The synonyms in the traffic information can be found by searching the definitions of synonyms of respective concepts in the concept table, and then the synonymous information processing device 103 transforms these synonyms into the concepts defined in the knowledge base. As such, for the traffic information which may contain words of different description forms but the same semantic, the synonyms in the traffic information are normalized to maintain unified description.
At S604, the checking device 104 checks whether every piece of traffic information is legal and valid so as to ensure the accuracy of the traffic information. The checking 104 performs the following checking on the traffic information
1. Existence checking: checking whether the locations, directions and traffic status contained in the traffic information have been defined in the concept table of the knowledge base, and if there is any undefined concept, determining by the checking device 104 whether any error occurs in the traffic information or the concept definition in the concept table of the knowledge base is incomplete.
The detailed determination approach is as follows. The checking device 104 searches the knowledge base, and misspelling may occur in the traffic information if the concept table of the knowledge base includes any concept similar to the undefined one; if the undefined concept has been appeared several times in the history traffic database, this concept may be valid, and the concept definition in the concept table of the knowledge base may not be complete. Alternatively, this determination can be done by an administrator. As an example, “Xueyuan” in “camera, Xueyuan, congested, 07-3-15 07:56” has no definition, and there are similar concepts “Xueyuan Road” and “Xueyuan Bridge” in the knowledge base, “Xueyuan” may be thus misspelled.
2. Completeness checking. A complete piece of traffic information must contain location, traffic status and time, and it may also contain information source and direction. The checking device 104 will discard the traffic information if it is incomplete. For example, “camera, Xueyuan Bridge, 07-3-15 07:56” is not complete since it lacks the description of traffic status.
3. Semantic error checking. The checking device 104 checks whether any semantic error is included in the traffic information based on the knowledge base. Checking whether any semantic error is present comprises:
checking whether the relation among several geographic entities in the traffic information is correct, based on the spatial relationship between geographic entities as defined by the relation section in the knowledge base; as an example, “user report, Lianxiang Bridge on the Fourth North Ring Road, congested, 07-3-15 07:56” is wrong semantically since Lianxiang Bridge lies on the Third North Ring Road;
checking whether the geographic directions are consistent with the geographic entities in the traffic information, based on the relationship between geographic entities and geographic directions as defined by the relation section in the knowledge base; as an example, “camera, Xueyuan Road, from west to east, congested, 07-3-15 07:56” is wrong semantically since the direction of Xueyuan Road is from south to north.
4. Time expiration checking. The update speed of traffic information is very quick, and it is thus necessary to check whether the traffic information transmitted by each information source has been expired. Determining the expiration of traffic information is fulfilled by comparing the time value in the traffic information with the current time. If the difference between the time value included in the traffic information and the current time of the apparatus exceeds a predetermined threshold, it is determined that the traffic information has been expired.
5. Redundancy checking. The heterogeneous information sources may transmit repetitive traffic information, and it is necessary to delete the repeated and redundant information.
Further, the reliability of each information source can be determined and adjusted with reference to the history statistics of information checking. For example, if the information source “user report” frequently transmits some illegal traffic information, it is determined that this information source has a low reliability, and thus a lower value can be set for the reliability of this source.
At S605, the knowledge base mapping device 105 normalizes the traffic information and maps it to the knowledge base.
At S606, the traffic information integrating device 106 integrates the traffic information inconsistent in space, time and semantic based on the knowledge base 108.
At S607, the output device 107 outputs the integrated traffic information.
While the gathering and processing of traffic information are illustrated by example of Chinese traffic information, it is apparent in the art that the present invention can be applied to the traffic information gathering and processing in any other language, such as English and Japanese.
Next, a description will be given to the knowledge base mapping method in conjunction with
At S902, the transforming unit 1052 transforms the indirect information in the traffic information into traffic points based on the knowledge base 108. The indirect information means that the geographic entities described in the traffic information are actually not traffic points (e.g., a bridge or an intersection) but geographic objects near these traffic points. Therefore, it is necessary to transform the indirect information into the traffic information about some traffic points. For example, the indirection information “Beihang West Gate” contained in the traffic information “user report, Beihang West Gate, from north to south, unblocked, 07-3-15 09:12” is not a traffic point but a geographic object near the traffic point “Xuezhi Bridge”. The transforming unit 1052 determines whether any indirect information is included in the traffic information based on the relation section defined in the knowledge base 108, and, if there is indirect information, transforms it into a specific traffic point according to a transforming rule on the basis of the knowledge base. The transforming rule can be set in the axiom section of the knowledge base according to the spatial relationship in the knowledge base or can be set in a memory. The transforming rule is represented as “isa(x, z) & R(x, y)→transform-to (y)”, which denotes that “x will be transformed into y if x belongs to geographic category z and has spatial relationship R with traffic point y”. As such, by using the transforming rule “isa(x, geographic object) & nearest-bridge (x, y)→transform-to(y)”, the indirection information “Beihang West Gate” contained in the traffic information “user report, Beihang West Gate, from north to south, unblocked, 07-3-15 09:12” can be transformed into its nearest point “Xuezhi Bridge”.
Based on the decomposition and transformation of the traffic information, the association between respective pieces of traffic information can be found out so as to facilitate the traffic information integration by the traffic information integrating device 106.
Next, at S903, the mapping unit 1053 maps all the traffic information to the electronic map in accordance with the longitude and latitude coordinates of geographic entities as well as the spatial relationship between geographic entities defined in the attribute and relation sections of the knowledge base.
At S904, the knowledge base mapping device 105 outputs the mapped traffic information to the traffic information integrating device 106.
It should be noted that this knowledge base mapping device 105 is illustrated merely as an example. The mapping unit 1053 can be omitted so that the knowledge base mapping device 105 can output the combination of decomposed and transformed traffic information directly to the traffic information integrating device 106. Alternatively, the knowledge base mapping device 105 can also include only the decomposing unit 1051 or the transforming unit 1052.
Referring to
At S1201, the traffic information is inputted from the knowledge base mapping device 105.
At S1202, the spatial consistency processing unit 1061 integrates the spatially inconsistent traffic information based on the knowledge base 108 to obtain a spatially consistent result. The spatial conflict processing part 10611 and the spatial complementary processing part 10612 in the spatial consistency processing unit 1061 conduct spatial conflict processing and spatial complementary processing on the traffic information, respectively, so as to acquire spatially-related traffic information which is consistent in space.
In particular, the spatial conflict processing carried out by the spatial conflict processing part 10611 comprises correcting the contradict traffic description information in several pieces of spatially-related traffic information. If a location has a different traffic description from those of locations spatially related to the location (i.e., locations adjoining the location) at adjoining moments, the spatial conflict processing part 10611 corrects such traffic description information as follows.
Spatial conflict detection: first, extracting locations contained in the respective pieces of traffic information, and finding out the adjoining relationship between these locations by referring to the spatial relationship defined in the relation section of the knowledge base; then, for each of these locations, comparing its traffic with those of the adjoining locations at adjoining moments, and confirming the occurrence of spatial conflict if the traffic of the location is different from those of most of the adjoining locations.
Spatial conflict elimination: correcting the traffic information of the location having spatial conflict to be consistent with the traffic information of most of the adjoining locations. Alternatively, it can be determined whether to perform the spatial conflict elimination in accordance with reliability of information sources as well as comparison with history traffic data. For example, if the current traffic information about a location is inconsistent with that of several adjacent locations, the traffic information originates from an unreliable information source (e.g., user report), and the traffic information is not consistent with its history traffic data, either, it is necessary to correct the traffic information so as to maintain consistence with the traffic information about the adjacent locations.
a) shows an example of the spatial conflict processing on the traffic information. As can be seen from the spatial relationships “between (Xueyuan Bridge, Baofushi Bridge, Jianxiang Bridge)” and “between (Xuezhi Bridge, Xueyuan Bridge, Jimen Bridge)”, the locations adjacent to the Xueyuan Bridge include Baofushi Bridge, Jianxiang Bridge, Xuezhi Bridge and Jimen Bridge. The traffic status (unblocked) of Xueyuan Bridge from user report is inconsistent with the traffic status (congested) of the adjacent traffic point (Baofushi Bridge, Jianxiang Bridge, Xuezhi Bridge), and the traffic information comes from a relatively unreliable information source “user report”. Further, it can be derived from the history traffic data that the history traffic of Xueyuan Bridge at this moment is mainly congested. Therefore, the spatial consistency processing unit 1061 corrects the traffic status in the traffic information about “Xueyuan Bridge” to be “congested”.
The spatial complementary processing conducted by the spatial complementary processing part 10612 includes correcting incomplete traffic description information in several pieces of spatially-related traffic information. For a location with no traffic information provided, its traffic information can be deduced from that of the adjoining locations (spatially-related locations). The specific steps by the spatial complementary processing part 10612 comprises:
deducing based on rules: matching the traffic information transmitted from respective information sources with a rule library established in advance in conjunction with the spatial relationship defined in the relation section of the knowledge base; finding a matched rule, and generating new traffic information based on the rule. Alternatively, the history traffic database can be search simultaneously, and the newly-generated traffic information must be discarded if it is not consistent with the history traffic information at this moment or if it has existed.
The rule library is configured in advance based on the spatial relationship defined in the knowledge base. For example, the rule library can be set in the axiom section of the knowledge base or stored in a memory. A spatial complementary rule library can be preset in accordance with the spatial relationship defined in the knowledge base. For example, the complementary rule “between (z, x, y) & near (x, y) & equal (traffic (x), traffic (y))→equal (traffic (z), traffic (x))” means “it can be deduced that the traffic of the location z is the same as the location x if z lies between x and y, x is near y, and the traffic of x and y is the same at adjacent moments”.
b) shows an example of spatial complementary processing, in which the traffic status of Xueyuan Bridge and Jimen Bridge is the same (congested) at 17:58, 07-3-15, the two bridges are not far from each other. Since Xuezhi Bridge is located between them, and most of history traffic status of Xuezhi Bridge at this time is congested, a new piece of traffic information can be complemented as “Xuezhi Bridge, from north to south, congested, 07-3-15 17:58”.
At S1203, the time consistency processing unit 1062 integrates the temporally inconsistent traffic information. The time conflict processing part 10621 and the time complementary processing part 10622 in the time consistency processing unit 1062 conduct time conflict processing and time complementary processing on the traffic information, respectively, so as to acquire the traffic information which is consistent in time.
The time conflict processing part 10621 in the time consistency processing unit 1062 processes the traffic information of which the description is inconsistent for the same location at different moments based on the knowledge base 108. The processing steps by the time conflict processing part 10621 comprise:
1) clustering, that is, clustering the traffic information about the same location: first, clustering the input traffic information in terms of location, gathering the traffic information about the same location into the same category, each category containing the traffic information of the same location at different moments; alternatively, plotting a traffic data graph, that is, for each category, automatically plotting a traffic data graph based on the traffic statuses of the location at different moments, with time being x-axis and traffic status being y-axis; the traffic data graph depicts the traffic status of a certain location changing with time.
As an alternative, besides the clustering of the input traffic information directly in accordance with location, the history traffic database can be referred to, and the traffic information of each location at the previously adjacent time can be classified into the category regarding this location.
2) conflict seeking, that is, seeking temporally-inconsistent traffic information: for each category, finding a coordinate point having a traffic status inconsistent with that at adjacent moments in the traffic data graph, the corresponding traffic information being time conflict information.
3) time conflict elimination, that is, eliminating the time conflict traffic information: correct such traffic information to keep consistent with the traffic information at adjacent moments. Alternatively, the processing on temporally-inconsistent information can be done simultaneously in accordance with reliability of information sources as well as comparison with history traffic data. If the temporally-inconsistent information originates from an unreliable information source, or it is not consistent with its history traffic data at the same moment, there must be some error in the traffic information, and it is necessary to correct the traffic information so as to maintain consistence with the traffic information at adjacent moments.
a) shows an example of the time conflict processing. For the location “Xuezhi Bridge” in the traffic information, its traffic statuses are congested at 17:58, 18:00 and 18:01, while a piece of traffic information from user report is that the traffic status is unblocked at 17:59. Since the traffic status of Xuezhi Bridge is congested in most cases as can be seen from the history traffic data, and the information source “user report” has a low reliability, the traffic status at 17:59 is correct as congested.
The time complementary processing part 10622 of the time consistency processing unit 1062 handles, based on the knowledge base 108, the situation where the traffic information of a location at certain time is absent. The steps by the time consistency processing unit 1062 comprise:
1) clustering, that is, clustering the traffic information about the same location: first, clustering the input traffic information in terms of location, gathering the traffic information about the same location into the same category, each category containing the traffic information of the same location at different moments; alternatively, plotting a traffic data graph, that is, for each category, automatically plotting a traffic data graph based on the traffic statuses of the location at different moments, with time being x-axis and traffic status being y-axis; the traffic data graph depicts the traffic status of a certain location changing with time.
As an alternative, besides the clustering of the input traffic information directly in accordance with location, the history traffic database can be referred to, and the traffic information of each location at the previously adjacent time can be classified into the category regarding this location.
2) absence information seeking, that is, seeking the traffic information absent at certain time point or the traffic information lacking traffic status: for each category, finding the absent coordinate points in the traffic data graph.
3) deducing: analyzing in the traffic data graph the traffic statuses of the coordinate points adjacent to each absent coordinate point, adding new traffic information or complementing the traffic information lacking traffic status if the traffic information of these adjacent coordinate points is substantially identical, with the location in the added or complemented traffic information being the location corresponding to one of the adjacent coordinate points, the time being the moment corresponding to the absent coordinate point, and the traffic status being that of the adjacent coordinate points.
As an alternative, the history traffic database can be referred to, and the newly-added traffic information must be discarded if it is not consistent with the history traffic information at this moment or if it has existed.
b) shows an example of time complementary processing. For the location “Xueyuan Bridge” in the traffic information, its traffic statuses are congested at 8:20, 8:21 and 8:23, while its traffic information at 8:22 is not present. Since the traffic statuses at adjacent moments are congested, a piece of traffic information is added as “integrating result, Xueyuan Bride, from west to east, congested, 07-3-15 8:22”.
At S1204, the semantic consistency processing unit 1063 processes the traffic information, which is about the same location but inconsistent in traffic status, into semantically consistent traffic information based on the knowledge base 108. The semantic consistency processing unit 1063 integrates the semantically inconsistent traffic information by correcting the traffic information which is inconsistent regarding the same location. The semantic conflict processing part 10631 and the semantic complementary processing part 10632 conduct semantic conflict processing and semantic complementary processing on the traffic information, respectively, so as to acquire traffic information which is consistent in semantics.
The semantic conflict processing part 10631 finds out the semantically conflicting traffic information and then eliminates such semantic conflict through the following steps:
1) semantic conflict determination, that is, determining whether there is semantic conflict between at least two pieces of traffic information. For the same location, different information sources provide contradict traffic descriptions at adjacent moments, while only one description is proper. The approach is to determine this according to specific semantic relationship between different traffic statuses defined in the relation section of the knowledge base 108. As an example, referring to the relation table shown in
2) semantic conflict elimination: determining the reliability of semantically conflicting traffic information, and then retaining the most reliable information. The reliability determination for traffic information can be done in accordance with the following rules:
reliability of information source, that is, selecting the traffic information transmitted from a reliable information source. The reliability of an information source is calculated from the provider, data update speed and history information. For example, the provider of information source “user report” is ordinary user, the update speed is low, and such source often transmits some illegal or inaccurate traffic information. So, the reliability of “user report” is low, and the reliability of the traffic information from it is accordingly low.
Time comparison, that is, the nearer the time value in the traffic information is with respect to the current moment of the apparatus, the more reliable the traffic information is.
Majority having priority, that is, if different information sources provide different descriptions about the traffic status of the same location at adjacent moments, the majority has priority. For example, if most information sources consider a location as congested at the moment, only a few sources reports that this location has no congested, the traffic status of congestion is more reliable.
Comparison with history traffic data, that is, comparing the traffic information with the history traffic data of the same location at this moment. The more consistent the two pieces of information are, the more reliable the traffic information is.
The semantic complementary processing part 10632 finds the traffic information semantically complementary to each other and then combine such information in a semantic sense. The involved steps comprise:
1) semantic complementary determination, that is, determining whether there is semantic complement between at least two pieces of traffic information. The occurrence of semantic complement is characterized in that, for the same location, different information sources provide traffic descriptions which are different but complementary to each other at adjacent moments, and these descriptions are all proper. The approach is to determine according to specific semantic relationship between different traffic statues defined in the relation section of the knowledge base 108. As an example, referring to the relation table shown in
2) semantic complementary combination, that is, for the traffic information semantically complementary to each other, combining the traffic statuses based on specific semantic relationship between the traffic statuses. Combination rules can be pre-established according to various semantic relationships. For example, “causal (x, y)→generate (“x resulting from y”)” denotes that, if y is the reason for x, x and y can be combined as “x resulting from y”. Since causal (congested, traffic accident), “congested” and “traffic accident” can be combined as “congested resulting from traffic accident”. The combination rules can be set in the axiom section of the knowledge base.
With the traffic information processing apparatus and method of the present invention, it is possible to integrate effectively the traffic information from a variety of heterogeneous information sources and ensure the accuracy, completeness and reliability of traffic information.
Although the present invention has been illustrated above with reference to the detailed embodiments, the present invention is not limited to the described embodiments and defined only by the appended claims. It will be understood that any modification and change made to the embodiments by those skilled in the art within the scope and spirit of the present invention.
| Number | Date | Country | Kind |
|---|---|---|---|
| 2007 1 0127120 | Jun 2007 | CN | national |
| Number | Name | Date | Kind |
|---|---|---|---|
| 5164904 | Sumner | Nov 1992 | A |
| 5173691 | Sumner | Dec 1992 | A |
| 5523950 | Peterson | Jun 1996 | A |
| 5933100 | Golding | Aug 1999 | A |
| 6278965 | Glass et al. | Aug 2001 | B1 |
| 6356836 | Adolph | Mar 2002 | B1 |
| 6944642 | Hopmann et al. | Sep 2005 | B1 |
| 6965574 | Cook et al. | Nov 2005 | B1 |
| 7266443 | Hirose | Sep 2007 | B2 |
| 7512489 | Endo et al. | Mar 2009 | B2 |
| 20040225433 | Burt | Nov 2004 | A1 |
| 20070094384 | Matsumura et al. | Apr 2007 | A1 |
| 20070118275 | Qi et al. | May 2007 | A1 |
| 20070118564 | Qi et al. | May 2007 | A1 |
| Number | Date | Country |
|---|---|---|
| 1967524 | May 2007 | CN |
| 2002-260162 | Sep 2002 | JP |
| 2006-23886 | Jan 2006 | JP |
| 2007-73024 | Mar 2007 | JP |
| 2007-141243 | Jun 2007 | JP |
| Number | Date | Country | |
|---|---|---|---|
| 20090037087 A1 | Feb 2009 | US |
