Search Tool FESS

Search Tool FESS

Previous section, I introduced the search technology in general. It is thought that a search system can be built by assembling a database and a query DSL and UI for it, and it is possible to build them from scratch by combining several open sources, but verification experiments with unclear required specifications. In the case of, an agile approach is required to first create a working module and build it while receiving feedback from the user. In order to do that, it is a good idea to use a framework that can be constructed as easily as possible.

Here are some of the tools you can use for them. First of all, FESS. It is an open source full-text search tool and is touted as “a full-text search server that can be easily built in 5 minutes.” Since it works with Java, it works on all environments of Windows, Mac, and Lynux, and it is provided under the Apache license, so it can be used for free.

The UI screen of FESS is shown below.

In addition to general full-text search, the search function includes such as location information search linked with map information, facet drilldown for user dictionary registration, and classification by labeling search results. 

Also, regarding the corresponding data, file crawl for document (MS Office, PDF, etc.) files, web crawl for specified web pages, or data store crawl for DB is possible, and automatic data import and indexing are possible.

The installation / configuration of the tool is explained in detail on the FESS website with screenshots, but here is a simple procedure as follows.

(1) First check the Java version. For mac, enter the following in the terminal, and if there is a message like the one below, JAVA is included. In this case the version is 11.0 (JAVA 11 is recommended for FESS),

> java --version
java 11.0.1 2018-10-16 LTS
Java(TM) SE Runtime Environment 18.9 (build 11.0.1+13-LTS)
Java HotSpot(TM) 64-Bit Server VM 18.9 (build 11.0.1+13-LTS, mixed mode)

(2) If JAVA is not included (or the version is old), install JAVA.

(3) Go back to (1), check the installed JAVA version, and then download FESS. (Just download and unzip the latest fess zip file on the page)

(4) Double-click the FESS file in the bin folder from the FESS folder downloaded in (3) (or specify the FESS folder in the terminal and execute / bin / fess.

(5) By (4), FESS will start up, and if you access http: // localhost: 8080 / admin with a browser, you will enter the FESS management screen.

(6) Define the verification purpose, and determine the file, web page, and DB to be input. In the case of files, the search target is one file, so if you want to search by page or paragraph, create a file that divides them into the required units.

(7) The file of (6) can be stored in any folder on the PC, and the data can be imported from the crawl setting page, but for easy operation, select the simple setup of the management screen.

First, log in at http://localhost:8080/admin page. Enter the username “admin” and the password “admin” as follows

Next is the first screen after logging in. Press the “System” button.

Then press “Wizard” under “System”.

Press the “Start Configuration” button.

Set the crawl settings (Name: arbitrary, Crawl path: absolute path of the file folder stored in (6), Max access: number of target data + α, Depth: folder structure in the stored data, if any, or 1 if not) and press the “Create” button.(Note that file paths are expressed differently in Windows.)

Press Start Crawl.

(8) Go to “Scheduler” on the Administration page and check the processing status. When all statuses are “Enabled,” all settings are complete. Click “Search Screen” in the upper right corner of the screen.

Then the search screen appears.

Enter the keywords, and when the results come up, you are done.

(9) If you want to tune the search conditions, refer to the search page on the official website and tune.

Business issues often start with “searching”, and by using FESS, a search system can be easily constructed, and verification experiments can be conducted in the shortest possible time. After rotating the experiment to some extent, you can think about the next step from the analyzed problems, use them if the problems are likely to be solved by using FESS as it is, and if more advanced problems appear, FESS You will consider using Elastic Saerch, which is used in the background of, and also consider a custom-made system.

Next time, I will talk about Elasticsearch, which can build a search system with higher flexibility than FESS.