Browse Prior Art Database

Parameterized Error Log Information Search

IP.com Disclosure Number: IPCOM000230070D
Publication Date: 2013-Aug-16
Document File: 4 page(s) / 208K

Publishing Venue

The IP.com Prior Art Database

Abstract

Disclosed are a method and system by which a search engine considers the customized objects in an error log (e.g., table, view, stored procedure, dimension, etc.) as parameters as a means of finding the best results and possible resolutions to errors.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 50% of the total text.

Page 01 of 4

Parameterized Error Log Information Search

While searching for resolution to a software error, users often enter the error log information as a query and perform an Internet search. The error log contains some customized objects (e.g., table name, view name, stored procedure, class, etc.). Due to this customization, appropriate search results are difficult to reach.

As an example of the problem, when the user searches for the error log, "object vw_test_sample", it is not found in database.

Figure 1: User does not find a result with the above search query

If the user modifies the search query by removing the object name, then the system returns more results.

Figure 2: Removing "vw_test_sample" from the search query returns multiple results

1


Page 02 of 4

Currently, to produce good results, the user must modify the error log information and then use the same as a search query. If the log information is complex, then while removing the object name the user might miss an important keyword, which may return

inappropriate results. This approach is time consuming, as the user has to modify the search query.

Herein disclosed is a method and system by which a search engine considers the customized objects (e.g., table, view, stored procedure, dimension, etc.) as parameters as a means of finding the best results. While searching the parameterized search query, the system identifies a parameter (based on pattern) and accordingly displays appropriate search results. The user does not have to filter any error log information. For training purpose, the user can provide additional information such as the type of the customized object (e.g., table, view, stored procedure, dimension etc.). In addition,

while posting any resolution, the user can provide the object type. Otherwise, automatic training is done based on text mining. The system presents the user with a category of object names (i.e., type of object), and the user selects the appropriate category to reach the desired results.

Figure 3 illustrates how the user can define the parameter type while searching any such content. According to the example, the user has defined "vw_test_sample" as "View Name". These data are for training, and the search provider uses the parameter type for training. Once a good amount of data is stored, and search engine is training, then the user does not have to provide "parameter type" again.

Similarly, while posting or replaying any comments, the user has to provide the parameter type. Again, this information is used for training purpose.

Figure 3: User can define the parameter type while searching

Figure 4 illustrates how a user can get category based search results for the same search query. In this case, using the error log as search query (without any changes)

2


Page 03 of 4

returns the search results with a different category. The user can select any category and can get search results in that category.

Figure 4: A user can get category based search resu...