7.20.2. Completion¶

This section describes about the following completion features:

How it works
How to use
How to learn

7.20.2.1. How it works¶

The completion feature uses three searches to compute completed words:

Prefix RK search against registered words.

Cooccurrence search against learned data.

Prefix search against registered words. (optional)

7.20.2.1.1. Prefix RK search¶

See Prefix RK search for prefix RK search.

If you create dataset which is named as query by groonga-suggest-create-dataset executable file, you can update pairs of registered word and its reading by loading data to _key and kana column of item_query table explicitly for prefix RK search.

7.20.2.1.2. Cooccurrence search¶

Cooccurrence search can find registered words from user’s partial input. It uses user input sequences that will be learned from query logs, access logs and so on.

For example, there is the following user input sequence:

input	submit
s	no
se	no
sea	no
sear	no
searc	no
search	yes
e	no
en	no
eng	no
engi	no
engin	no
engine	no
enginen	no (typo!)
engine	yes

Groonga creates the following completion pairs:

input	completed word
s	search
se	search
sea	search
sear	search
searc	search
e	engine
en	engine
eng	engine
engi	engine
engin	engine
engine	engine
enginen	engine

All user not-submitted inputs (e.g. “s”, “se” and so on) before each an user submission maps to the submitted input (e.g. “search”).

To be precise, this description isn’t correct because it omits about time stamp. Groonga doesn’t case about “all user not-submitted inputs before each an user submission”. Groonga just case about “all user not-submitted inputs within a minute from an user submission before each an user submission”. Groonga doesn’t treat user inputs before a minute ago.

If an user inputs “sea” and cooccurrence search returns “search” because “sea” is in input column and corresponding completed word column value is “search”.

7.20.2.1.3. Prefix search¶

Prefix search can find registered word that start with user’s input. This search doesn’t care about romaji, katakana and hiragana not like prefix RK search.

This search isn’t always ran. It’s just ran when it’s requested explicitly or both prefix RK search and cooccurrence search return nothing.

For example, there is a registered word “search”. An user can find “search” by “s”, “se”, “sea”, “sear”, “searc” and “search”.

7.20.2.2. How to use¶

Groonga provides suggest command to use completion. --type complete option requests completion.

For example, here is an command to get completion results by “en”:

Execution example:

suggest --table item_query --column kana --types complete --frequency_threshold 1 --query en
# [
#   [
#     0,
#     1337566253.89858,
#     0.000355720520019531
#   ],
#   {
#     "complete": [
#       [
#         1
#       ],
#       [
#         [
#           "_key",
#           "ShortText"
#         ],
#         [
#           "_score",
#           "Int32"
#         ]
#       ],
#       [
#         "engine",
#         1
#       ]
#     ]
#   }
# ]

7.20.2.3. How it learns¶

Cooccurrence search uses learned data. They are based on query logs, access logs and so on. To create learned data, Groonga needs user input sequence with time stamp and user submit input with time stamp.

For example, an user wants to search by “engine”. The user inputs the query with the following sequence:

2011-08-10T13:33:23+09:00: e

2011-08-10T13:33:23+09:00: en

2011-08-10T13:33:24+09:00: eng

2011-08-10T13:33:24+09:00: engi

2011-08-10T13:33:24+09:00: engin

2011-08-10T13:33:25+09:00: engine (submit!)

Groonga can be learned from the input sequence by the following command:

load --table event_query --each 'suggest_preparer(_id, type, item, sequence, time, pair_query)'
[
{"sequence": "1", "time": 1312950803.86057, "item": "e"},
{"sequence": "1", "time": 1312950803.96857, "item": "en"},
{"sequence": "1", "time": 1312950804.26057, "item": "eng"},
{"sequence": "1", "time": 1312950804.56057, "item": "engi"},
{"sequence": "1", "time": 1312950804.76057, "item": "engin"},
{"sequence": "1", "time": 1312950805.86057, "item": "engine", "type": "submit"}
]

7.20.2.4. How to update reading data¶

Groonga requires registered word and its reading for prefix RK search. This section describes how to register a word and its reading.

Here is an example to register “日本” which means Japan in English:

Execution example:

load --table event_query --each 'suggest_preparer(_id, type, item, sequence, time, pair_query)'
[
{"sequence": "1", "time": 1312950805.86058, "item": "日本", "type": "submit"}
]
# [[0, 1337566253.89858, 0.000355720520019531], 1]

Here is an example to update reading data to complete “日本”:

Execution example:

load --table item_query
[
{"_key":"日本", "kana":["ニホン", "ニッポン"]}
]
# [[0, 1337566253.89858, 0.000355720520019531], 1]

Then you can complete registered word “日本” by Romaji input - “nihon”.

Execution example:

suggest --table item_query --column kana --types complete --frequency_threshold 1 --query nihon
# [
#   [
#     0,
#     1337566253.89858,
#     0.000355720520019531
#   ],
#   {
#     "complete": [
#       [
#         1
#       ],
#       [
#         [
#           "_key",
#           "ShortText"
#         ],
#         [
#           "_score",
#           "Int32"
#         ]
#       ],
#       [
#         "日本",
#         2
#       ]
#     ]
#   }
# ]

Without loading above reading data, you can’t complete registered word “日本” by query - “nihon”.

You can register multiple readings for a registered word because kana column in item_query table is defined as a Vector column.

This is the reason that you can also complete the registered word “日本” by query - “nippon”.

Execution example:

suggest --table item_query --column kana --types complete --frequency_threshold 1 --query nippon
# [
#   [
#     0,
#     1337566253.89858,
#     0.000355720520019531
#   ],
#   {
#     "complete": [
#       [
#         1
#       ],
#       [
#         [
#           "_key",
#           "ShortText"
#         ],
#         [
#           "_score",
#           "Int32"
#         ]
#       ],
#       [
#         "日本",
#         2
#       ]
#     ]
#   }
# ]

This feature is very convenient because you can search registered word even though Japanese input method is disabled.

If there are multiple candidates as completed result, you can customize priority to set the value of boost column in item_query table.

Here is an example to customize priority for prefix RK search:

Execution example:

load --table event_query --each 'suggest_preparer(_id, type, item, sequence, time, pair_query)'
[
{"sequence": "1", "time": 1312950805.86059, "item": "日本語", "type": "submit"}
{"sequence": "1", "time": 1312950805.86060, "item": "日本人", "type": "submit"}
]
# [[0, 1337566253.89858, 0.000355720520019531], 2]
load --table item_query
[
{"_key":"日本語", "kana":"ニホンゴ"}
{"_key":"日本人", "kana":"ニホンジン"}
]
# [[0, 1337566253.89858, 0.000355720520019531], 2]
suggest --table item_query --column kana --types complete --frequency_threshold 1 --query nihon
# [
#   [
#     0,
#     1337566253.89858,
#     0.000355720520019531
#   ],
#   {
#     "complete": [
#       [
#         3
#       ],
#       [
#         [
#           "_key",
#           "ShortText"
#         ],
#         [
#           "_score",
#           "Int32"
#         ]
#       ],
#       [
#         "日本",
#         2
#       ],
#       [
#         "日本人",
#         2
#       ],
#       [
#         "日本語",
#         2
#       ]
#     ]
#   }
# ]
load --table item_query
[
{"_key":"日本人", "boost": 100},
]
# [[0, 1337566253.89858, 0.000355720520019531], 1]
suggest --table item_query --column kana --types complete --frequency_threshold 1 --query nihon
# [
#   [
#     0,
#     1337566253.89858,
#     0.000355720520019531
#   ],
#   {
#     "complete": [
#       [
#         3
#       ],
#       [
#         [
#           "_key",
#           "ShortText"
#         ],
#         [
#           "_score",
#           "Int32"
#         ]
#       ],
#       [
#         "日本人",
#         102
#       ],
#       [
#         "日本",
#         2
#       ],
#       [
#         "日本語",
#         2
#       ]
#     ]
#   }
# ]

7.20.1. Introduction

7.20.3. Correction