KR101282501B1 - Computer readable record medium where broadcast stream with index is recorded, apparatus for generating broadcast stream with index and method of generating broadcast stream with index - Google Patents

Abstract

In order to improve energy efficiency and access time efficiency, a broadcast stream structure having a new index structure for indexing heterogeneous XML documents from an XML document set having a heterogeneous schema, and an apparatus and method for generating a broadcast stream having a new index structure are provided. present. The proposed broadcast stream structure is a broadcast stream structure generated from a set of heterogeneous XML documents, and includes an XML document field containing contents of heterogeneous XML documents, and an address value satisfying a location path and a location path for heterogeneous XML documents. It includes an index field including, and the index field is placed in front of the XML document field. By placing a position path for a heterogeneous XML document and an index including an address value satisfying the position path at the front of the broadcast stream in one broadcast period or by placing a plurality of indexes, the access time to the target document for a given query is improved. It is shortened and energy efficiency is improved.

Description

Computer readable record medium where broadcast stream with index is recorded, apparatus for generating broadcast stream with index and method of generating broadcast stream with index}

The present invention relates to a broadcast stream structure having an index, an apparatus and a method for generating a broadcast stream having an index, and more particularly, to a broadcast stream structure having an index capable of broadcasting XML data in a wireless environment, and An apparatus and method for generating a broadcast stream having an index.

EXtensible Markup Language (XML) documents have components such as Elements, Tags, Attributes, Entities, PCDATA, and CDATA.

The following is a brief description of each component.

Elements are key components in both XML and HTML documents. To make the elements easier to understand, look at the elements in the HTML document. In an HTML document, the "body" and "table" tags are examples of elements. In the XML document example we have seen so far, "note", "to", etc. are also elements. An element can contain text, another element, or it can be empty. Examples of empty HTML elements include "hr", "br", and "img".

Tags are used to construct elements. A start tag such as <element name> indicates the beginning of an element, and a end tag such as </ element name> indicates the end of an element.

Attributes provide additional information about an element. The attribute is placed within the start tag of the element. Attributes appear as name / value pairs. The following example is for an attribute that provides additional information about the source file in the img element:

<img src = "computer.gif" />

In this case, the element name is img. The name of the attribute is src. The value of the property is computer.gif. This element ends with / because it is an empty element.

PCDATA stands for parsed character data. Character data is the text placed between the start and end tags of an XML element. PCDATA is the text to be interpreted by the parser. Tags in the text will be treated as markup and entities will be expanded. In simpler terms, this means that tags such as <and> will be parsed by the XML parser, and the entity will also be parsed and referenced by the value pointed to by the entity.

CDATA stands for character data. CDATA is text that won't be interpreted by the parser. In fact, all text in the XML document is interpreted by the XML parser, but is output as it is, not as much as CDATA. Tags within text will not be treated as markup and entities will not be expanded. This can easily be interpreted to mean that tags and entities such as <and> will not be interpreted by the XML parser, so characters and entities such as <and> will be output as is.

Entities can be thought of as variables. An entity is a unit of character set that can be referenced within a document. For example, if you have a phrase that you use repeatedly, you can easily represent it using an entity. Entities include regular entities and parameter (parameter) entities. References to defined entities are called entity references.

One conventional XML data broadcasting method is an S-Node. In the S-Node, when XML data is converted into stream data for broadcasting, duplicate structure information in the original XML data is left as it is.

1 illustrates an environment in which a plurality of mobile clients that receive data broadcast with a server exist. The server 10 broadcasts various data including XML data in a stream. Mobile clients 12, such as laptops, cell phones, PDAs, etc., receive the broadcasted data stream 14 to satisfy the query.

FIG. 2 is a diagram schematically showing an example of original XML data, showing XML data in a tree form. To get the data you want from XML data, you use XQuery, the world wide consortium's standardized query language.

For example, the result of a simple XPath query such as / dblp / article / title / for the XML data of FIG. 2 is the text of nodes 3, 14, and 17 of FIG. 2. The values ABC, HI, and JKL. At this time, to get the result value corresponding to the query, all paths having the path / dblp / article / title / must be searched.

Thus, in the conventional XML stream generation and reception method, the server 10 continues to broadcast the data stream 14, and the mobile client 12 in need thereof receives the necessary data from the broadcast channel.

In this case, the waiting time for the mobile client 12 to obtain the desired data from the data stream 14 becomes long, which is pointed out as a disadvantage in terms of latency-efficiency.

In addition, the mobile client is a portable device having limited battery resources, there is a problem that it is difficult to conserve energy by receiving unnecessary data.

The present invention has been proposed to solve the above-described problems, and has a broadcast stream structure having a new index structure for indexing heterogeneous XML documents from an XML document set having heterogeneous schemas to improve energy efficiency and access time efficiency. The purpose is to provide.

Another object of the present invention is to provide an apparatus and method for generating a broadcast stream having the new index structure.

In order to achieve the above object, a broadcast stream structure having an index according to a preferred embodiment of the present invention is a broadcast stream structure generated from a set of heterogeneous XML documents, and an XML document containing contents of heterogeneous XML documents. A field, and an index field including a location path for a heterogeneous XML document and an address value satisfying the location path, wherein the index field is disposed in front of the XML document field.

The address value means the address of the bucket that contains the first XML node among all the response nodes that satisfy the location path.

The address of the bucket indicates the time when the bucket is broadcast in the wireless broadcast channel.

The index field further includes identification information of heterogeneous XML documents.

At least one index field is disposed in one broadcast period.

If there are two index fields in one broadcast period, the second index field includes a location path for heterogeneous XML documents carried in the next broadcast period following one broadcast period and an address value satisfying the location path. .

Broadcast stream generation apparatus having an index according to a preferred embodiment of the present invention, the data stream generation unit for generating a data stream for the received heterogeneous XML document; A location identifier which identifies a location of XML elements having different location paths in the data stream; An index entry generation unit for generating an index entry including a location path for a heterogeneous XML document and an address value satisfying the location path based on the information from the location identification unit; And an index generator for generating an index that is a set of generated index entries.

The data stream generator connects the generated data stream to a data stream for XML documents previously generated.

The index generator places the index in front of the field containing the content of the heterogeneous XML document in the broadcast stream including the field containing the content of the heterogeneous XML document.

The address value means the address of the bucket that contains the first XML node among all the response nodes that satisfy the location path.

The address of the bucket indicates the time when the bucket is broadcast in the wireless broadcast channel.

The index entry generating unit adds identification information of heterogeneous XML documents to the index entry.

The index generator arranges at least one index within one broadcast period.

According to a preferred embodiment of the present invention, a method of generating a broadcast stream having an index includes: generating a data stream for a received heterogeneous XML document; Identifying locations of XML elements having different location paths in the data stream; Generating an index entry including a location path for a heterogeneous XML document and an address value satisfying the location path based on the location identified information; And generating an index that is a set of generated index entries.

The data stream generation step connects the generated data stream to the data stream for the XML documents previously generated.

The index generation step places the index in front of the field containing the content of the heterogeneous XML document in the broadcast stream including the field containing the content of the heterogeneous XML document.

The address value means the address of the bucket that contains the first XML node among all the response nodes that satisfy the location path.

The address of the bucket indicates the time when the bucket is broadcast in the wireless broadcast channel.

The index entry generation step adds identification information of heterogeneous XML documents to the index entry.

The index generation step arranges at least one index within one broadcast period.

According to the present invention having such a configuration, by placing a plurality of indexes including position paths for heterogeneous XML documents and address values satisfying the position paths in front of the broadcast stream in one broadcast period or by placing a plurality of indexes, The access time to the target document is shortened and the energy efficiency is improved.

1 is a diagram illustrating an environment in which a server and a plurality of mobile clients exist.
2 is a diagram illustrating original XML data in a tree form.
3 is a diagram illustrating a broadcast stream structure having an index according to an embodiment of the present invention.
4 is a block diagram of an apparatus for generating a broadcast stream having an index according to an embodiment of the present invention.
5 is a flowchart illustrating a method of generating a broadcast stream having an index according to an embodiment of the present invention.
6 is a flowchart illustrating an operation of receiving and querying a broadcast stream according to an embodiment of the present invention on the mobile client side.
7 is a view for explaining a modification of the broadcast stream structure having an index according to an embodiment of the present invention.
8 is a diagram illustrating a difference between a conventional broadcast stream and a broadcast stream structure having an index according to an embodiment of the present invention.

Hereinafter, a broadcast stream structure having an index, a broadcast stream generating apparatus having an index, and a method according to an embodiment of the present invention will be described with reference to the accompanying drawings. Prior to the detailed description of the present invention, terms and words used in the present specification and claims should not be construed as limited to ordinary or dictionary terms. Therefore, the embodiments described in this specification and the configurations shown in the drawings are merely the most preferred embodiments of the present invention and do not represent all the technical ideas of the present invention. Therefore, It is to be understood that equivalents and modifications are possible.

3 is a diagram illustrating a broadcast stream structure having an index according to an embodiment of the present invention.

A broadcast stream with an index according to an embodiment of the invention is generated from a set of XML documents with heterogeneous schemas.

The broadcast stream 20 of the embodiment of the present invention includes an index field 22 containing a new type of index called a prefixsummary and an XML document field 24a, 24b, and 24c containing contents of heterogeneous XML documents. , 24d). In FIG. 3, the index field 22 is disposed first in one broadcast cycle, and the XML document fields 24a, 24b, 24c, and 24d are subsequently disposed. In other words, the index field 22 is attached to the front of the XML data stream. Although FIG. 3 shows four XML document fields, the number may be added or subtracted.

The structure of the index carried in the index field 22 will be described below. First, when a prefix summation having a depth h is PS (h), PS (h) is a set of index entries p and a. Here, p means the location path of length h extracted from the given heterogeneous XML documents. a means the address of the bucket containing the XML node that appears first among all the response nodes that satisfy the location path p in the broadcast stream. In this way, the structure of the index is represented by PS (h), which is a set of index entries (p, a).

The length of the location path means the number of location steps included in the path. In the embodiment of the present invention, it is assumed that the location paths to be considered for the indexes start from the root element in the XML document. In addition, the address of the bucket indicates the time (or delivery time of the bucket) when the bucket is broadcast in the wireless broadcast channel.

The index shown in the index field 22 of FIG. 3 includes a first index entry indicating identification information (DocID) of heterogeneous XML documents, a second index entry indicating a location path for heterogeneous XML documents, and a position. And a third index entry representing an address value Ari. Satisfying the path.

Figure 3 shows a broadcast stream with a prefix summation of depth two. The index contains link addresses for different location paths corresponding to length 2 derived from the XML data stream. For example, the index entry corresponding to the location path "/ S / Z" contains the address t ₆ for the bucket whose element satisfying the location path first appears in the XML data stream.

4 is a block diagram of an apparatus for generating a broadcast stream having an index according to an embodiment of the present invention. It is understood that a broadcast stream generating device having an index is provided in a server.

The broadcast stream generating apparatus of FIG. 4 includes a receiver 30, a data stream generator 32, a position identifier 34, an index entry generator 36, and an index generator 38.

The receiver 30 receives a plurality of heterogeneous XML documents.

The data stream generator 32 generates a data stream using a streaming technique σ for each heterogeneous XML document received by the receiver 30. Here, the streaming technique may use a simple text listing method or a commonly used streaming method. On the other hand, the streaming technique may be an existing XML streaming method using distributed indexes such as S-Node or DIX. The data stream generator 32 connects the generated data stream to a data stream for previously generated XML documents.

The position identifier 34 identifies (detects) positions of XML elements having different position paths in the XML data stream generated by the data stream generator 32.

The index entry generator 36 generates an index entry including a location path for the heterogeneous XML document and an address value satisfying the location path based on the information from the location identifier 34.

The index generator 38 generates an index that is a set of index entries generated by the index entry generator 36. That is, the index generator 38 generates the index (PS (h) of FIG. 3) based on the index entry extracted from the entire XML data stream. In arranging the generated index, the index generator 38 arranges the generated index in front of the field containing the content of the heterogeneous XML document in the broadcast stream including the field containing the content of the heterogeneous XML document. The index generator 38 may arrange at least one index within one broadcast period.

5 is a flowchart illustrating a method of generating a broadcast stream having an index according to an embodiment of the present invention.

First, the receiver 30 receives a plurality of heterogeneous XML documents (S10). A plurality of heterogeneous XML documents received are delivered to the data stream generator 32.

The data stream generator 32 generates a data stream using a streaming technique for each heterogeneous XML document from the receiver 30 (S12). The streaming technique in the data stream generator 32 may be a simple text listing method or an existing XML streaming method using distributed indexes such as S-Node or DIX.

The data stream generator 32 connects the generated data stream to a data stream for previously generated XML documents (S14).

Thereafter, the position identifier 34 identifies (detects) positions of XML elements having different position paths in the XML data stream generated by the data stream generator 32 (S16).

The index entry generation unit 36 generates an index entry including a location path for the heterogeneous XML document and an address value satisfying the location path based on the information identified by the location identification unit 34 (S18). Here, the location path refers to the access path from the data stream to the location where the content of the XML document is loaded, and the address value is a value of the bucket including the XML node that appears first among all response nodes satisfying the location path in the data stream. It means an address (that is, a time (or delivery time of the bucket) when the bucket is broadcast in the wireless broadcast channel).

The generated index entry is transferred to the index generator 38.

Accordingly, the index generator 38 generates an index that is a set of index entries from the index entry generator 36 (S20). That is, the index generator 38 generates an index represented by PS (h) based on the index entry extracted from the entire XML data stream.

Thereafter, the index generator 38 attaches the generated index to the field containing the content of the heterogeneous XML document in the broadcast stream including the field containing the content of the heterogeneous XML document. The index generator 38 outputs the broadcast stream with the index as the final broadcast stream (S22).

Table 1 below illustrates a broadcast stream generation algorithm.

INPUT : a set D of XML documents, path length h
OUTPUT : XML broadcast stream S
1 IndexEntries : = φ ; DataStream : = φ ;
2 FOR-EACH (XML document d in D ) {
3 Generate broadcast stream Sd of d using a streaming method σ;
4 DataStream : = DataStream ∥ Sd ;
5 Paths : = φ ;
6 FOR-EACH (element E in Sd in the order of appearance in Sd ) {
7 p: = prefix of length h of the location path specification of E ;
8 IF ( p is not in paths ) {
9 a: = address of the bucket in DataStream which contains E ;
10 IndexEntries : = IndexEntries U (p, a);
11 Paths : = Paths U p;
12}
13}
14}
15 Construct PS (h) from IndexEntries ;
16 Place PS (h) and DataStream into a broadcast stream S ;
17 RETURN S ;

In Table 1, for each XML document, a broadcast stream S _d is generated using a specific streaming scheme (σ), and the generated broadcast stream is connected to the streams for previously generated XML documents ( See lines 3-4). For the XML data stream generated as shown in lines 6-12, identify the locations of XML elements with different location paths and create a prefix summation index entry of the form (location path, address value for the first result node). do. Finally, based on the index entry extracted from the entire XML data stream, a prefix summation index structure is created and attached to the beginning of the XML stream (see lines 15-16).

6 is a flowchart illustrating an operation of receiving and querying a broadcast stream according to an embodiment of the present invention on the mobile client side.

In the following, it is assumed that a mobile client uses a query for a simple location path. In addition, the mobile client is considered to have a built-in module that stores a program capable of receiving and querying a broadcast stream using a prefixsummary. It is assumed that the protocol is matched between the mobile client and the server.

The mobile client searches for the index (prefix sumhead) closest to the initial reception position of the broadcast stream broadcast from the server (S30). In this case, it is assumed that each bucket includes an address PSLink indicating the location of the nearest index and a type indicating the type of the bucket. Therefore, the mobile client can receive the first bucket of the index at any time, regardless of when the stream is received.

Thereafter, the mobile client searches for an entry corresponding to a location path corresponding to the given query (S32).

If the entry exists, the reception of another bucket is omitted using the address value included in the entry and the bucket is directly received (S34).

Then, the first node that satisfies the location path in the associated XML document is received (S36).

Thereafter, the mobile client continues to search the selected XML document to search all the result sets satisfying the location path (S38). Accordingly, only the results of the query in the stream can be selectively received.

Table 2 below illustrates an algorithm for a mobile client to receive and query a broadcast stream using a prefixsummary.

INPUT : a user query in the form of a location path p
OUTPUT : a set of XML nodes satisfying p
1 Result : = φ ;
2 Read the current bucket B on the broadcast channel;
3 IF ( B.Type ! = PS-BUCKET) {
4 Read the next nearest PrefixSummary using B.PSLink ;
5}
6 Find the index entry IE in PrefixSummary that matches a prefix of p ;
// Find the first answer node a ₁ in the stream based on IE.address
7 a ₁ : = getFirstAnswer (IE.address) ;
8 Result : = Result U a ₁ ;
9 a _n : = a ₁ ;
10 WHILE (there exists a remaining answer in the stream) {
// Find the next answer node a _{n + 1} in the stream;
11 a _{n + 1} : = getNextAnswer (a _n );
12 Result : = Result U _{an + 1} ;
13 a _n : = a _{n + 1} ;
14}
15 RETURN Result ;

In the case of Table 2, the mobile user first searches for the PrefixSummary closest to the broadcast stream first reception position (see lines 2 to 5). At this time, it is assumed that each bucket includes an address (PSLink) indicating the location of the nearest PrefixSummary and a TYPE indicating the type of the bucket. Therefore, the mobile user can receive the nearest PrefixSummary regardless of the stream reception time. Next, the user retrieves the entry corresponding to the location path p that matches the given query (see line 6). If there is an entry (that is, an entry that satisfies p), the reception of another bucket is skipped by using the address value included in the entry and the bucket is directly received. Receive the first node that satisfies the location path p in the associated XML document using the getFirstAnswer () operation described in line 7. The mobile user then continues to search the selected XML document to search all the result sets that satisfy p (see lines 9-14). Here, the getNextAnswer () operation described in line 11 efficiently retrieves the next result node using the approach used in the XML stream structure in the stream. For example, in the S-Node scheme, the indexes included in the S-nodes, that is, the index address for the next node with the same path as the same-tag sibling link, which is the index address for the next node with the same name, are same- By using the path link, only the result values of the query in the stream can be selectively received.

7 is a view for explaining a modification of the broadcast stream structure having an index according to an embodiment of the present invention.

In FIG. 7, the access time to the index is shortened by replicating the index several times. Thus, the mobile client has improved access time efficiency. In the modified example of the present invention, the (1, M) index replication technique is used in consideration of efficiency. The (1, M) index duplication scheme allows the index information to appear in every 1 / M portion of the entire data by repeatedly arranging the index information in one broadcasting period. This (1, M) index replication scheme is used in the index generator 38. That is, the index generator 38 arranges at least one index within one broadcast period. When the (1, M) index replication scheme is used, the broadcast stream generation algorithm described in Table 1 above must be modified, but the query processing algorithm of Table 2 does not need to be modified.

7 shows an example in which PrefixSummary PS 2 is duplicated and arranged in one broadcast period. When the PrefixSummary cloned in FIG. 7 is PS ₂ , the address value of the index entry corresponding to the location paths "/ a / b", "/ a / c", "/ r / a", and "/ r / s" Is different from the original PrefixSummary PS ₁ . This is because the address values for these paths correspond to the next broadcast period, so the time values delivered are different.

As shown in FIG. 7, if the mobile user starts to receive a broadcast stream for query processing (tune-in point) on the stream of document d ₁ or d ₂ , and the target document for the query is on d ₄ . When present, the mobile user can find the first result of d ₄ (that is, the response node) that exists in the same period using the next PS ₂ index entry without having to wait until PS ₁ of the next broadcast period. Therefore, access time to the data stream is shortened by using replicated indexes. In other words, latency can be reduced in query processing.

In FIG. 7, it is assumed that a copy of PrefixSummary is stored in the next broadcast period when a copy of PrefixSummary is located in a stream of a specific document such as PS ₂ existing in d ₃ to simplify index generation and query processing. .

FIG. 8 is a diagram illustrating a difference between a conventional broadcast stream and a broadcast stream structure having an index according to an embodiment of the present invention. FIG. 8 (a) shows a conventional streaming method without a prefixsummary. FIG. 8B illustrates a query evaluation process on a stream of heterogeneous XML documents generated by FIG. 8B. FIG. 8B illustrates a query evaluation process on a broadcast stream having a prefix summation.

In the case of Fig. 8A, the mobile client first accesses the starting bucket of the XML document close to the first tune-in point (i.e. probe wait) on the stream to access the target data on the broadcast stream. Find. After reaching the start bucket of the XML document, the mobile client finds the relevant XML document that can respond to the given user query by examining the start bucket of subsequent XML documents sent over the air.

As such, FIG. 8A sequentially searches the starting buckets of various documents existing in front of the target document to find the desired target document in order to access the desired target document.

In this case, if the desired target document is at the end, the arrival time to the target document is considerably longer by finding the starting bucket of all XML documents.

However, when using the index (PrefixSummary) including the location path and the address value satisfying the location path for heterogeneous XML documents as shown in FIG. It will skip the bucket reception and go directly to the desired target document. Accordingly, it is possible to quickly access a desired target document as compared with FIG. 8A.

While the present invention has been described in connection with what is presently considered to be practical exemplary embodiments, it is to be understood that the invention is not limited to the disclosed embodiments, but, on the contrary, is intended to cover various modifications and equivalent arrangements included within the spirit and scope of the appended claims. You must see.

30: receiver 32: data stream generator
34: location identification unit 36: index entry generation unit
38: index generator

Claims (12)

A computer-readable recording medium that records a broadcast stream generated from a set of heterogeneous XML documents.
The broadcast stream includes an XML document field containing contents of the heterogeneous XML document, and an index field including a location path for the heterogeneous XML document and an address value satisfying the location path.
The index field is placed in front of the XML document field,
Two or more index fields are included in one broadcast period, and the second or more index fields correspond to a location path for a heterogeneous XML document carried in the next broadcast period following the one broadcast period and an address satisfying the location path. A computer-readable recording medium having recorded therein a broadcast stream having an index, the structure comprising a value. The method according to claim 1,
And the address value is an address of a bucket including the first XML node among all the response nodes satisfying the location path. delete delete A data stream generator for generating a data stream for the received heterogeneous XML document;
A location identifier which identifies a location of XML elements having different location paths in the data stream;
An index entry generator for generating an index entry including a location path for the heterogeneous XML document and an address value satisfying the location path based on the information from the location identifier; And
And an index generator for generating an index that is a set of the generated index entries.
The index entry generating unit includes two or more indexes in one broadcast period, and the second or more indexes indicate a location path and a location path for heterogeneous XML documents carried in a next broadcast period following the one broadcast period. A broadcast stream generating device having an index, characterized by including an address value that satisfies. The method according to claim 5,
And the data stream generator is configured to connect the generated data stream to a data stream for previously generated XML documents. The method according to claim 5,
And the index generator is configured to place the index in front of the field containing the content of the heterogeneous XML document in the broadcast stream including the field containing the content of the heterogeneous XML document. . delete Generating a data stream for the received heterogeneous XML document;
Identifying locations of XML elements having different location paths in the data stream;
Generating an index entry including a location path for the heterogeneous XML document and an address value satisfying the location path based on the location identified information; And
Generating an index that is a set of the generated index entries;
The generating of the index entry includes two or more of the indexes within one broadcast period, and the second or more indexes include location paths for heterogeneous XML documents carried in the next broadcast period following the one broadcast period, and And generating an address value satisfying the location path. The method according to claim 9,
The data stream generating step is a broadcast stream generation method having an index, characterized in that for linking the generated data stream to the data stream for the previously generated XML documents. The method according to claim 9,
The index generation step includes generating a broadcast stream having an index, wherein the index is placed in front of a field containing the content of the heterogeneous XML document in a broadcast stream including the field containing the content of the heterogeneous XML document. Way. delete

Images