RankingAgorithm is a search library that uses a new scoring algorithm to rank results accurately and relevantly. RankingAlgorithm is very easy to use since it uses the Apache Lucene** index but ranks and scores on its own.
Multiple algorithms are available, SIMPLE* and COMPLEX*. SIMPLE is a very fast algorithm and can return queries in <10ms on a 10m wikipedia index (complete index).
It can also scale to 2 billion docs. COMPLEX is a more complex algorithm so is a little
slower compared to the SIMPLE, but can also still return queries in < 10ms on a 10m wikipedia index (complete index).
COMPLEX is more accurate and should be able to give you the best rankings as compared to SIMPLE. SIMPLE is also very good and may be well suited than COMPLEX for some type of queries.
RankingAlgorithm works in two modes, document mode and product mode:
In document mode, it ranks documents such as HTML, Wikipedia, Word/PDF docs relevantly while in Product mode, a term's occurrence is taken into account and scored accordingly. So titles starting with "wii console" are ranked first, and the others rank lower as the occurrence of "wii console" shifts in the title or gets reversed, see below:
- Wii Console and Wii Fit Plus with Balance Board Bundle (Nintendo Wii)
- Wii Console System with Wii Sports Resort Game with TWO MotionPlus Attachments
- Nintendo Wii Console w/ Bonus Wii Sports Resort Bundle, Black
- Pelican Accessories Wii Console Stand - Nintendo Wii
- Grafitti Skin for Nintendo Wii Console
- NEW AC Adapter Cable Cord Power Supply For NINTENDO WII Gaming Console
- Wii Remote Charging Console Stand
- Nintendo Wii Skin - System Console Skin and two Wii Remote Skins - Blue Vortex
- CET Domain 10301901 Console Stand Station for Nintendo Wii
RankingAlgorithm is very easy to use as Apache Lucene the very popular open source library can be used create the index. The results from Solr (again a very popular open source search platform) with RankingAlgorithm seems to be comparable to Google and much better than Lucene for Perl index searches, see comparison.
Features
1. Search very accurate and relevant. Comparable to Google site search and much better than Lucene, see comparison.
2. Two algorithms, SIMPLE and COMPLEX.
3. Two modes, Document and Product mode. Product mode enables very accurate product/retail/short twitter text searches. Document mode enables relevant search, can be used for product searches too.
4. By default AND/OR combinations.
5. Very easy scoring with a relevancy score.
6. Very easy to use.
7. Score boosting, supports Document, Field, Query & Query term boosts.
8. Uses the very popular Lucene index. No changes to your code or index.
9. NRT search
10. Query a 10m wikipedia index in <10 ms, scale upto 2 billion docs
11. Built in AutoComplete components for autocomplete similar google/yahoo/bing
12. Supports Glob/Regular expressions/Fuzyy/Prefix/Suffix queries, results ranked relevantly
Search components
Autocomplete using Solr with RankingAlgorithm, similar to Google/Yahoo/Bing's autocomplete (It is free),
Give it a try
Autocomplete components bundled with RankingAlgorithm (very fast), similar to Google/Yahoo/Bing's autocomplete (It is free),
Give it a try
Demo
Documentation
Browse the javadocs here:
RankingAlgorithm 1.4.7 docs,
RankingAlgorithm 1.4.7 Javadocs.
RankingAlgorithm 1.4.6 docs,
RankingAlgorithm 1.4.6 Javadocs.
RankingAlgorithm 1.4.5 docs,
RankingAlgorithm 1.4.5 Javadocs.
RankingAlgorithm 1.4.4 docs,
RankingAlgorithm 1.4.4 Javadocs.
RankingAlgorithm 1.4.3 docs,
RankingAlgorithm 1.4.3 Javadocs.
RankingAlgorithm 1.4.1 docs,
RankingAlgorithm 1.4.1 Javadocs.
RankingAlgorithm 1.3 docs,
RankingAlgorithm 1.3 Javadocs.
RankingAlgorithm 1.2 docs,
RankingAlgorithm 1.2 Javadocs.
RankingAlgorithm 1.1/1.0 docs,
RankingAlgorithm 1.1 Javadocs.
WIKI,
Apache Solr with RankingAlgorithm WIKI.
Browse RankingAlgorithm usage
Downloads (Give it a try) (It is Free))
- Download RankingAlgorithm Library (Free)
- ver 1.4.8, works with Lucene 4.3 [md5sum:a40e448358657de69291ee1c517d3a64], 2013/06/92 (1.4.8.20130602.1)
(Has built in components Edge/Infix autocomplete with source. Support for glob/regular expressions in boolean queries)
- ver 1.4.7, works with Lucene 4.1/4.0 [md5sum:ea6d5ca2c727026a05778827], 2013/03/18 (1.4.7.20130318.1)
(H/as built in components Edge/Infix autocomplete with source. Support for glob/regular expressions in boolean queries)
- ver 1.4.6, works with Lucene 4.x [md5sum:61cab6301d30fa681ab198cf57355e42], 2012/12/23 (1.4.5.20121105.6)
(Has built in components Edge/Infix autocomplete with source. Support for glob/regular expressions in boolean queries)
- ver 1.4.5, works with Lucene 4.x [md5sum:97ff962f5bb1a060591f35e07effa8fd], 2012/12/09 (1.4.5.20121105.4)
(Has built in components Edge/Infix autocomplete with source. Support for glob/regular expressions in boolean queries)
- ver 1.4.4, works with Lucene 4.x [md5sum:adcd42f3d3f869a4aff315e735bbbe5d], 2012/10/29 (1.4.4.20121029.1)
(Has built in components Edge/Infix autocomplete with source)
- ver 1.4.4-alpha, works with Lucene 4.x [md5sum:3529ec4251b8c4f99fc0a1873a889940], 2012/10/12 (1.4.4.2012012.1)
(Has built in components Edge/Infix autocomplete with source)
- ver 1.4.3, works with Lucene 3.x [md5sum:b148a3b6a2fe7650d0d522e9fecd9035], 2012/08/01 (1.4.3.20120801.1)
(Has built in components Edge/Infix autocomplete with source)
- ver 1.4.1, works with Lucene 3.x [md5sum:93954336e081f97c8d60f9aabb99bead], 2012/03/05 (1.4.1.20120305.1)
- ver 1.3, works with Lucene 3.x [md5sum:c3b04c45771c3df557e2d0e1236a3bf3], 2011/11/25 (1.3.20111125.1)
- ver 1.2, works with Lucene 3.x [md5sum:4ef61cbef033a567f434d7008ae808fe], 2011/08/21
- ver 1.1, works with Lucene 3.x [md5sum:119e6f17466d245fe857e9c427ba88dd]
- ver 1.0, works with Lucene 3.x [md5sum:8896655a465b4ee2b447bdacc7ad404e]
- Download RankingAlgorithm exposed as Lucene API Library (Free)
Steps:
Lucene 4.0
1. Download Lucene core 4.x.0 above
2. Download RankingAlgorithm 1.4.1 for Lucene 4.x above
3. Replace lucene-core-4.x.0.jar with lucene-core-40-alpha.jar
3a. unzip lucene-core-4.x.0.zip
3b. mv lucene-core-4.x.0.jar /backup/lucene-core-4.x.0.jar
3c. cp lucene-core-40-alpha.jar lucene-core-4.x.0-ALPHA.jar
4. Add lucene-core-4.x.0-ALPHA.jar and rankingalgorithm40_1.4.4.jar to the classpath
5. Run your previous Lucene examples, applications with no changes but using the RankingAlgorithm for searches instead of Lucene library.
Lucene 3.0
1. Download Lucene core 3.x.0 above
2. Download RankingAlgorithm 1.3 for Lucene 3.x above
3. Replace lucene-core-3.x.0.jar with lucene-ra-core-3.x.0.jar
3a. unzip lucene-core-3.x.0.zip
3b. mv lucene-core-3.x.0.jar lucene-core-3.x.0.jar_backup
3c. cp lucene-ra-core-3.x.0.jar lucene-core-3.x.0.jar
4. Add RankingAlgorithm to the classpath
5. Run your previous Lucene examples, applications with no changes but using the RankingAlgorithm for searches instead of Lucene library.
Note:
a. To use product mode, start your app with -Dmode=product, default is document mode. To use document mode, use -Dmode=document
b. To specify the scan depth in product mode, use -Dscan=fast, -Dscan=medium, -Dscan=full
c. To specify AND/AND_OR/OR, specify -DAND_OR=and, -DAND_OR=or, -DAND_OR=and_or
d. To switch to lucene, start with -Dlucene=true
- Download Apache Solr with RankingAlgorithm (Free)
Bundle:
- Solr 4.3 (Realtime-search support) with RankingAlgorithm 1.4.8 (bundle 165MB)
[very high update perf(70000 docs / sec), very high performance, <10 ms response times (billion docs), scale upto 2 billion docs; Realtime-search allows you to search in realtime while indexing without a commit, more info SolrJ based Realtime-search Concurrent Update Testcode]
[md5sum_43_148.txt], 2013/04/11 (4.3.20130527.1)
- Solr 4.21 (Realtime-search support) with RankingAlgorithm 1.4.8 (bundle 165MB)
[very high update perf(70000 docs / sec), very high performance, <10 ms response times (billion docs), scale upto 2 billion docs; Realtime-search allows you to search in realtime while indexing without a commit, more info SolrJ based Realtime-search Concurrent Update Testcode]
[md5sum_148.txt], 2013/04/11 (4.21.20130411.1)
- Solr 4.1 (Realtime-search support) with RankingAlgorithm 1.4.7 (bundle 142MB)
[very high update perf(70000 docs / sec), very high performance, <10 ms response times (billion docs), scale upto 2 billion docs; Realtime-search allows you to search in realtime while indexing without a commit, more info SolrJ based Realtime-search Concurrent Update Testcode]
[md5sum_147.txt], 2013/05/06 (4.1.20130211.3)
- Solr 4.0 (Realtime-search support) with RankingAlgorithm 1.4.6 (bundle 142MB)
[very high update perf(70000 docs / sec), very high performance, <10 ms response times (billion docs), scale upto 2 billion docs; Realtime-search allows you to search in realtime while indexing without a commit, more info SolrJ based Realtime-search Concurrent Update Testcode]
[md5sum:4964210928a12d882084664e13f9454f], 2012/12/23 (4.0.20121022.6)
- Solr 3.6.2 (NRT support) with RankingAlgorithm 1.4.3 (bundle 118MB) -- (includes TimedSerialMergeScheduler)
[very high update perf(5000 docs in 450 ms), 50-100% improved performance over 1.4, <50 ms response times (10 million wikipedia index), scale upto 100m docs, more info SolrJ NRT Testcode]
[md5sum:c6a8e421d38a3f4dbe44d94a26c4ba76], 2013/01/20 (3.6.2.20130120.1)
- Solr 3.6 (NRT support) with RankingAlgorithm 1.4.3 (bundle 118MB)
[very high update perf(5000 docs in 450 ms), 50-100% improved performance over 1.4, <50 ms response times (10 million wikipedia index), scale upto 100m docs, more info SolrJ NRT Testcode]
[md5sum:eb4347d1257805d496773e220dae1b33], 2012/07/29 (3.6.20120527.2)
- Solr 3.5 (NRT support) with RankingAlgorithm 1.4.2 (bundle 118MB)
[very high update perf(5000 docs in 450 ms), 50-100% improved performance over 1.4, <50 ms response times (10 million wikipedia index), scale upto 100m docs, more info SolrJ NRT Testcode]
[md5sum:334ab7fc30e3910e0976a45bb22cb8fc], 2012/04/18 (3.5.20111211.8)
- Solr 3.4 (NRT support) with RankingAlgorithm 1.3 (bundle 110MB)
[very high update perf(10000 docs/sec), 50-100% improved performance over 1.2, <50 ms response times (10 million wikipedia index), scale upto 100m docs, more info SolrJ NRT Testcode]
[md5sum:5bf96814ff1ddffe255804e49b447baf], 2011/11/25 (3.4.20111125.1)
- Solr 3.3 (NRT support) with RankingAlgorithm 1.2 (bundle 110MB)
[very high update perf(10000 docs/sec), <100 ms response times (10 million wikipedia index), scale upto 100m docs, more info SolrJ NRT Testcode]
[md5sum:6515d2da2071d0f6c76b5e44acfce63d], 2011/08/23 (3.3.20110823.1)
- Solr 3.2 (NRT support) with RankingAlgorithm 1.1 (bundle 110MB)
[high update perf, 5000 docs/sec] [md5sum:be03b95be21c788186b7bdc34fe38f67]
- Download Apache Solr 1.4.1 with RankingAlgorithm 1.1 (apache solr + rankingalgorithm bundle)
[md5sum:2c4b264c92ff9d32285048a4c780cf27]
(see installation steps)
War file:
- Solr 4.3 (Realtime-search support) with RankingAlgorithm 1.4.8 (war file 20MB)
[very high update perf(70000 docs / sec), very high performance, <10 ms response times (billion index), scale upto 2 billion docs; Realtime-search allows you to search in realtime while indexing without a commit more info SolrJ based Realtime-search Concurrent Update Testcode]
[md5sum_43_148.txt], 2013/05/27 (4.3.20130527.1)
- Solr 4.21 (Realtime-search support) with RankingAlgorithm 1.4.8 (war file 20MB)
[very high update perf(70000 docs / sec), very high performance, <10 ms response times (billion index), scale upto 2 billion docs; Realtime-search allows you to search in realtime while indexing without a commit more info SolrJ based Realtime-search Concurrent Update Testcode]
[md5sum_148.txt], 2013/04/11 (4.21.20130411.1)
- Solr 4.1 (Realtime-search support) with RankingAlgorithm 1.4.7 (war file 15MB)
[very high update perf(70000 docs / sec), very high performance, <10 ms response times (billion index), scale upto 2 billion docs; Realtime-search allows you to search in realtime while indexing without a commit more info SolrJ based Realtime-search Concurrent Update Testcode]
[md5sum:md5sum_147.txt], 2013/05/06 (4.1.20130211.3)
- Solr 4.0 (Realtime-search support) with RankingAlgorithm 1.4.6 (war file 15MB)
[very high update perf(70000 docs / sec), very high performance, <10 ms response times (billion index), scale upto 2 billion docs; Realtime-search allows you to search in realtime while indexing without a commit more info SolrJ based Realtime-search Concurrent Update Testcode]
[md5sum:0afd14a6e08a5fabadcfe23f2aa23d24], 2012/12/23 (4.0.20121022.6)
- Solr 3.6.2 (NRT support) with RankingAlgorithm 1.4.3 (war file 11MB) -- (includes TimedSerialMergeScheduler)
[very high update perf(5000 docs in 450 ms), 50-100% improved performance over 1.4, <50 ms response times (10 million wikipedia index), scale upto 100m docs, more info SolrJ NRT Testcode]
[md5sum:b4b574bad8805f80b42c88fd51272f69], 2013/01/20 (3.6.2.20130120.1)
- Solr 3.6 (NRT support) with RankingAlgorithm 1.4.3 (war file 11MB)
[very high update perf(5000 docs in 450 ms), 50-100% improved performance over 1.4, <50 ms response times (10 million wikipedia index), scale upto 100m docs, more info SolrJ NRT Testcode]
[md5sum:493672c0ccad1ce3393e38df23e23a16], 2012/07/29 (3.6.20120527.2)
- Solr 3.5 (NRT support) with RankingAlgorithm 1.4.2 (war file 6.64MB)
[very high update perf(5000 docs in 450 ms), 50-100% improved performance over 1.4, <50 ms response times (10 million wikipedia index), scale upto 100m docs, more info SolrJ NRT Testcode]
[md5sum:f527237639d8faa7cfa58bd77e14dd59], 2012/04/18 (3.5.20111211.8)
- Chunk download (1 MB each, last chunk maybe less than 1 MB):
On Unix/Linux/Cygwin/Mac: cat chunk01 chunk02 ... > get_orig_file_back.zip
On Windows: copy /B chunk01+chunk02 ... get_orig_file_back.zip
unzip get_orig_file_back.zip
Chunk 00
Chunk 01
Chunk 02
Chunk 03
Chunk 04
Chunk 05
Chunk 06
- Solr 3.4 (NRT support) with RankingAlgorithm 1.3 (war file 7MB)
[very high update perf(10000 docs/sec), 50-100% improved performance over 1.2, <50 ms response times (10 million wikipedia index), scale upto 100m docs, more info SolrJ NRT Testcode]
[md5sum:f5d572625774ae9a44a18c18bfc47898], 2011/10/11 (3.4.20111125.1)
- Solr 3.3 (NRT support) with RankingAlgorithm 1.2 (war file 7MB)
[very high update perf(10000 docs/sec), <100 ms response times (10 million wikipedia index), scale upto 100m docs, more info SolrJ NRT Testcode]
[md5sum:12e5462a0e3def665904e5a8914e435a], 2011/08/23 (3.3.20110823.1)
- Solr 3.2 (NRT support) with RankingAlgorithm 1.1 (war file 7MB)
[high update perf, 5000 docs/sec] [md5sum:85d0979a820912c34fa4ee70f642fc1d]
- Download Apache Solr 1.4.1 with RankingAlgorithm 1.1 (just the war file)
[md5sum:fec749de1fabc58bfe4863a7e8502ee8]
(see installation steps)
- Download Autocomplete using Solr with RankingAlgorithm (similar to Google/Yahoo/Bing's autocomplete)(Free)
Download elasticsearch with RankingAlgorithm (Free)
Install elasticsearch and RankingAlgorithm separately:
- Download elasticsearch from here as before. Install elasticsearch as in the instructions here.
- Follow below steps:
- 1. Download RankingAlgorithm exposed as lucene-core-3.6.1 below.
- 2. Download RankingAlgorithm 1.4.3
- 3. mv elasticsearch/lib/lucene-core-3.6.1.jar to /tmp
- 4. mv downloads/lucene-ra-core-3.6.1.jar elasticsearch/lib/lucene-core-3.6.1.jar
- 5. mv downloads/rankingalgorithm30_1.4.3 elasticsearch/lib
Discussion
RankingAlgorithm
** Apache Solr and Lucene are trademarks of Apache Software Foundation
