Datasets:
sent_id large_stringlengths 14 14 | tok_id int64 1 324 | word large_stringlengths 1 42 | predicate large_stringclasses 29
values | semantic_role large_stringlengths 0 124 |
|---|---|---|---|---|
ml_pos_0000001 | 1 | കന്യാസ്ത്രീയാകാൻ | P1 | P2:B-Content;P4:B-Theme |
ml_pos_0000001 | 2 | ആഗ്രഹിച്ച | P2 | P4:I-Theme |
ml_pos_0000001 | 3 | യുവതി | P2:B-Experiencer;P3:B-Theme;P4:I-Theme | |
ml_pos_0000001 | 4 | പിന്നീട് | P3:B-Temporal;P4:I-Theme | |
ml_pos_0000001 | 5 | പോൺ | P3:B-Result;P4:I-Theme | |
ml_pos_0000001 | 6 | താരമായി | P3:I-Result;P4:I-Theme | |
ml_pos_0000001 | 7 | മാറിയ | P3 | P4:I-Theme |
ml_pos_0000001 | 8 | സംഭവം | P4:I-Theme | |
ml_pos_0000001 | 9 | വാർത്തകളിൽ | P4:B-Location | |
ml_pos_0000001 | 10 | ഇടംപിടിച്ചു | P4 | |
ml_pos_0000001 | 11 | . | ||
ml_pos_0000002 | 1 | ഭാരത | P1:B-Location | |
ml_pos_0000002 | 2 | - | P1:I-Location | |
ml_pos_0000002 | 3 | പാകിസ്താൻ | P1:I-Location | |
ml_pos_0000002 | 4 | അതിർത്തിയിൽ | P1:I-Location | |
ml_pos_0000002 | 5 | വീണ്ടും | P1:B-Temporal | |
ml_pos_0000002 | 6 | വെടിവെപ്പുണ്ടാവുകയും | P1 | |
ml_pos_0000002 | 7 | ഒരു | P2:B-Patient | |
ml_pos_0000002 | 8 | ബി | P2:I-Patient | |
ml_pos_0000002 | 9 | . | P2:I-Patient | |
ml_pos_0000002 | 10 | എസ് | P2:I-Patient | |
ml_pos_0000002 | 11 | . | P2:I-Patient | |
ml_pos_0000002 | 12 | എഫ് | P2:I-Patient | |
ml_pos_0000002 | 13 | . | P2:I-Patient | |
ml_pos_0000002 | 14 | ജവാൻ | P2:I-Patient | |
ml_pos_0000002 | 15 | കൊല്ലപ്പെടുകയും | P2 | |
ml_pos_0000002 | 16 | ചെയ്തു | P2 | |
ml_pos_0000002 | 17 | . | ||
ml_pos_0000003 | 1 | എഴുന്നേറ്റുനിന്ന് | P1 | |
ml_pos_0000003 | 2 | കുനിഞ്ഞ് | P2 | |
ml_pos_0000003 | 3 | താഴെ | P3:B-Location;P4:B-Theme | |
ml_pos_0000003 | 4 | വെച്ചിരിക്കുന്ന | P3 | P4:I-Theme |
ml_pos_0000003 | 5 | വെള്ളം | P3:B-Theme;P4:I-Theme | |
ml_pos_0000003 | 6 | സ്ട്രോ | P4:B-Instrument | |
ml_pos_0000003 | 7 | ഉപയോഗിച്ച് | P4:I-Instrument | |
ml_pos_0000003 | 8 | കുടിക്കുന്നത് | P4 | |
ml_pos_0000003 | 9 | എക്കിൾ | P5:B-Theme | |
ml_pos_0000003 | 10 | മാറാനുള്ള | P5 | |
ml_pos_0000003 | 11 | ഏറ്റവും | ||
ml_pos_0000003 | 12 | ഫലപ്രദമായ | ||
ml_pos_0000003 | 13 | മാർഗമാണ് | ||
ml_pos_0000003 | 14 | . | ||
ml_pos_0000004 | 1 | എക്കിൾ | P1:B-Theme;P2:B-Stimulus;P3:B-Agent;P4:B-Agent | |
ml_pos_0000004 | 2 | മാറാൻ | P1 | P2:I-Stimulus;P3:I-Agent;P4:I-Agent |
ml_pos_0000004 | 3 | പ്രയാസം | P3:I-Agent;P4:I-Agent | |
ml_pos_0000004 | 4 | നേരിടുന്നവർ | P2 | P3:I-Agent;P4:I-Agent |
ml_pos_0000004 | 5 | തീർച്ചയായും | P3:B-Manner;P4:B-Manner | |
ml_pos_0000004 | 6 | ഈ | P3:B-Theme;P4:B-Theme | |
ml_pos_0000004 | 7 | രീതി | P3:I-Theme;P4:I-Theme | |
ml_pos_0000004 | 8 | പരീക്ഷിച്ചു | P3 | P4:B-Manner |
ml_pos_0000004 | 9 | നോക്കേണ്ടതാണ് | P4 | |
ml_pos_0000004 | 10 | . | ||
ml_pos_0000005 | 1 | ബലിയർപ്പണത്തിനായി | P4:B-Theme | |
ml_pos_0000005 | 2 | ബലിപീഠത്തിലെത്തുമ്പോൾ | ||
ml_pos_0000005 | 3 | മറ്റാരോടെങ്കിലും | P1:B-Recipient | |
ml_pos_0000005 | 4 | ക്ഷമിക്കാനുണ്ടെങ്കിൽ | P1 | |
ml_pos_0000005 | 5 | ബലിവസ്തുക്കൾ | P2:B-Theme | |
ml_pos_0000005 | 6 | അവിടെ | P2:B-Location | |
ml_pos_0000005 | 7 | സമർപ്പിച്ച് | P2 | |
ml_pos_0000005 | 8 | പശ്ചാത്തപിച്ചശേഷം | P3 | P4:B-Temporal |
ml_pos_0000005 | 9 | മാത്രമേ | P4:I-Temporal | |
ml_pos_0000005 | 10 | ബലി | P4:B-Theme | |
ml_pos_0000005 | 11 | അർപ്പിക്കാവൂ | P4 | |
ml_pos_0000005 | 12 | . | ||
ml_pos_0000006 | 1 | അല്ലാത്തപക്ഷം | ||
ml_pos_0000006 | 2 | ആ | P1:B-Theme | |
ml_pos_0000006 | 3 | ബലിയർപ്പണം | P1:I-Theme | |
ml_pos_0000006 | 4 | ദൈവം | P1:B-Agent | |
ml_pos_0000006 | 5 | സ്വീകരിക്കുന്നതല്ല | P1 | |
ml_pos_0000006 | 6 | . | ||
ml_pos_0000007 | 1 | ഒരു | P2:B-Manner | |
ml_pos_0000007 | 2 | വിശ്വാസി | P2:I-Manner | |
ml_pos_0000007 | 3 | എന്ന | P2:I-Manner | |
ml_pos_0000007 | 4 | നിലയിൽ | P2:I-Manner | |
ml_pos_0000007 | 5 | നൂറാവർത്തിയെങ്കിലും | P2:B-Degree | |
ml_pos_0000007 | 6 | വൈദികർ | P1:B-Agent;P2:B-Stimulus | |
ml_pos_0000007 | 7 | ബലിപീഠത്തിൽ | P1:B-Location;P2:I-Stimulus | |
ml_pos_0000007 | 8 | നിന്നുകൊണ്ട് | P1:I-Location;P2:I-Stimulus | |
ml_pos_0000007 | 9 | ഈ | P1:B-Theme;P2:I-Stimulus | |
ml_pos_0000007 | 10 | വചനം | P1:I-Theme | |
ml_pos_0000007 | 11 | ആഹ്വാനം | P1 | |
ml_pos_0000007 | 12 | ചെയ്യുന്നത് | P1 | |
ml_pos_0000007 | 13 | നാം | P2:B-Experiencer | |
ml_pos_0000007 | 14 | കേട്ടിട്ടുണ്ട് | P2 | |
ml_pos_0000007 | 15 | . | ||
ml_pos_0000008 | 1 | കോഴിക്കോട് | P2:B-Location | |
ml_pos_0000008 | 2 | ജില്ലയിലെ | P2:I-Location | |
ml_pos_0000008 | 3 | ആയഞ്ചേരിയിൽ | P2:I-Location | |
ml_pos_0000008 | 4 | സ്കൂൾ | P1:B-Theme;P2:B-Purpose | |
ml_pos_0000008 | 5 | ഉച്ചഭക്ഷണ | P1:I-Theme;P2:I-Purpose | |
ml_pos_0000008 | 6 | വിതരണം | P1:I-Theme;P2:I-Purpose | |
ml_pos_0000008 | 7 | കാര്യക്ഷമമാക്കാൻ | P1 | P2:I-Purpose |
ml_pos_0000008 | 8 | സർക്കാർ | P2:B-Agent | |
ml_pos_0000008 | 9 | നടപടികൾ | P2:B-Theme | |
ml_pos_0000008 | 10 | ആരംഭിച്ചു | P2 | |
ml_pos_0000008 | 11 | . | ||
ml_pos_0000009 | 1 | കഴിഞ്ഞ | ||
ml_pos_0000009 | 2 | ദിവസം | ||
ml_pos_0000009 | 3 | പൊതുവിദ്യാഭ്യാസ | P1:B-Location | |
ml_pos_0000009 | 4 | സെക്രട്ടറിയുടെ | P1:I-Location |
Malayalam Semantic Tags
A large semantic resource for Malayalam: 152,454 sentences (1,901,014 words) in which every predicate (the word or phrase expressing an event, action or state) is identified, and the words around it are labelled with their semantic role: who did it, to whom, where, when, how, why, with what.
Syntax tells you a word's grammatical form. Semantic roles tell you the meaning relation it holds to an event. In ഫിഫ ലോകകപ്പ് ട്രോഫി ഏപ്രിൽ ഏഴിന് അബുദാബിയിൽ പ്രദർശിപ്പിക്കും ("The FIFA World Cup trophy will be displayed in Abu Dhabi on April 7"), the predicate is പ്രദർശിപ്പിക്കും:
- what is displayed (Theme): the trophy;
- where (Location): അബുദാബിയിൽ;
- when (Temporal): ഏപ്രിൽ ഏഴിന്.
This who-did-what-to-whom structure is central to question answering, information extraction, event understanding, summarisation and machine translation.
| Sentences | 152,454 |
| Words | 1,901,014 |
| Predicate frames | 312,106 (about 2 per sentence) |
| Sentences with at least one predicate | 142,268 |
| Labelled arguments | 586,214 |
| Semantic roles | 22 |
The dataset comes in two views of the same annotation:
| Config | One row per | Use it for |
|---|---|---|
tokens (default) |
word | sequence labelling; aligning with other word-level resources |
frames |
predicate | predicate-argument extraction; event analysis |
from datasets import load_dataset
tokens = load_dataset("Navneeth017/malayalam-semantic-tags", "tokens", split="train")
frames = load_dataset("Navneeth017/malayalam-semantic-tags", "frames", split="train")
tokens view: columns
| Column | Type | Description |
|---|---|---|
sent_id |
string | Sentence identifier (ml_pos_0000001 …). |
tok_id |
int | Position of the word in its sentence, starting at 1. Sort by sent_id, tok_id to read sentences in order. |
word |
string | The word as it appears in the sentence. |
predicate |
string | If the word is (part of) a predicate, its predicate id within the sentence: P1, P2, … in left-to-right order. Empty otherwise. A predicate spanning several words (e.g. കൊല്ലപ്പെടുകയും ചെയ്തു) has the same id on each of its words. |
semantic_role |
string | The word's role(s) relative to the predicates of its sentence, as predicate_id:BIO-Role, separated by ; when the word belongs to arguments of more than one predicate. B- marks the first word of an argument and I- a continuing word. Empty if the word fills no role. |
Reading semantic_role: P2:B-Experiencer;P3:B-Theme;P4:I-Theme means this word:
- starts the Experiencer of predicate P2;
- starts the Theme of predicate P3;
- continues the Theme of predicate P4.
A word can play different roles for different events in the same sentence.
Example: കന്യാസ്ത്രീയാകാൻ ആഗ്രഹിച്ച യുവതി പിന്നീട് പോൺ താരമായി മാറിയ സംഭവം വാർത്തകളിൽ ഇടംപിടിച്ചു . ("The incident of the young woman who wished to become a nun later becoming a porn star made the news.")
| tok_id | word | predicate | semantic_role |
|---|---|---|---|
| 1 | കന്യാസ്ത്രീയാകാൻ | P1 | P2:B-Content;P4:B-Theme |
| 2 | ആഗ്രഹിച്ച | P2 | P4:I-Theme |
| 3 | യുവതി | P2:B-Experiencer;P3:B-Theme;P4:I-Theme | |
| 4 | പിന്നീട് | P3:B-Temporal;P4:I-Theme | |
| 5 | പോൺ | P3:B-Result;P4:I-Theme | |
| 6 | താരമായി | P3:I-Result;P4:I-Theme | |
| 7 | മാറിയ | P3 | P4:I-Theme |
| 8 | സംഭവം | P4:I-Theme | |
| 9 | വാർത്തകളിൽ | P4:B-Location | |
| 10 | ഇടംപിടിച്ചു | P4 | |
| 11 | . |
What each predicate expresses:
- P2 ആഗ്രഹിച്ച ("wished"): who wished = യുവതി (Experiencer); what = കന്യാസ്ത്രീയാകാൻ (Content).
- P3 മാറിയ ("became"): who = യുവതി (Theme); when = പിന്നീട് (Temporal); into what = പോൺ താരമായി (Result).
- P4 ഇടംപിടിച്ചു ("found a place"): what = the whole event phrase, words 1-8 (Theme); where = വാർത്തകളിൽ (Location).
frames view: columns
| Column | Type | Description |
|---|---|---|
sent_id |
string | Sentence identifier; matches the tokens view. |
predicate_id |
string | P1, P2, …; matches the predicate column of the tokens view. |
predicate |
string | The predicate's words. |
pred_start, pred_end |
int | tok_id of the predicate's first and last word (inclusive). |
arguments |
list | The predicate's arguments. Each has role, text, and start / end (tok_id of its first and last word, inclusive). |
Example row (predicate P3 above):
{"sent_id": "ml_pos_0000001", "predicate_id": "P3", "predicate": "മാറിയ", "pred_start": 7, "pred_end": 7,
"arguments": [{"role": "Theme", "text": "യുവതി", "start": 3, "end": 3},
{"role": "Temporal", "text": "പിന്നീട്", "start": 4, "end": 4},
{"role": "Result", "text": "പോൺ താരമായി", "start": 5, "end": 6}]}
Predicates
A predicate is the word or phrase at the centre of an event or state: usually a verb, in any of its Malayalam forms.
| Form | Examples |
|---|---|
| Finite verbs | തീരുമാനിച്ചു "decided", സ്ഥിരീകരിച്ചു "confirmed" |
| Participles | ആഗ്രഹിച്ച "who wished", പരിക്കേറ്റ "injured" |
| Infinitives and purpose forms | പിടികൂടാൻ "to catch" |
| Conditional forms | പറഞ്ഞാൽ "if said" |
| Nominalised verbs | ജീവിക്കുന്നത് "living" |
| Multi-word predicates (15,573) | രജിസ്റ്റർ ചെയ്ത "registered", കൊല്ലപ്പെടുകയും ചെയ്തു "was killed" |
Malayalam chains events through participles and converbs, so one sentence often contains several predicates (2 on average). Each one has its own set of arguments.
Semantic roles
There are 22 roles. Each example shows an argument [in brackets] and the predicate it belongs to (→).
Participants: who or what is involved
| Role | Count | Meaning | Example |
|---|---|---|---|
| Agent | 118,103 | The one who deliberately performs the action. | [ഞാൻ] → തീരുമാനിച്ചു: "I decided" |
| Theme | 183,625 | The thing or event that the action is about, moves, or is affected without change of state. The most frequent role. | അധികൃതർ ഉടൻ [നടപടി] → സ്വീകരിക്കണം: "authorities must take action" |
| Patient | 26,140 | The one that undergoes the action and is changed or affected by it. | [രോഗികളെ] മരണത്തിലേക്ക് → നയിക്കുമെന്ന: "will lead patients to death" |
| Experiencer | 10,562 | The one who feels, perceives or experiences something. | നട്ടെല്ലിന് പരിക്കേറ്റ [ഉദ്യോഗസ്ഥൻ]: "the officer who was injured" |
| Stimulus | 1,304 | What causes a feeling or perception. | [മാതൃകാപരമായ അദ്ദേഹത്തിന്റെ ജീവിതം] എല്ലാവരെയും → ആകർഷിച്ചു: "his exemplary life attracted everyone" |
| Content | 24,652 | What is said, thought, known or confirmed: the content of speech or cognition. | [രോഗബാധയില്ലെന്ന്] → സ്ഥിരീകരിച്ചു: "confirmed that there is no infection" |
| Recipient | 25,241 | The one who receives something or towards whom the action is directed. | [ശരീരത്തിന്] ദോഷം → ചെയ്തേക്കാം: "may harm the body" |
| Beneficiary | 1,780 | The one for whose benefit the action is done. | [വെസ്റ്റ് ഇൻഡീസിനായി] → കളത്തിലിറങ്ങി: "took the field for the West Indies" |
| Attribute | 8,623 | A property or status assigned to an entity. | പരിശോധനാഫലം [നെഗറ്റീവ്] → ആയിരുന്നു: "the test result was negative" |
| Topic | 1,403 | What something is about. | [ഒരു സിനിമയെക്കുറിച്ച്] അഭിപ്രായം → പ്രകടിപ്പിക്കാം: "can give an opinion about a film" |
Circumstances: where, when, how, why
| Role | Count | Meaning | Example |
|---|---|---|---|
| Location | 51,722 | Where the event takes place. | [അബുദാബിയിൽ] → പ്രദർശിപ്പിക്കും: "will be displayed in Abu Dhabi" |
| Source | 8,028 | Starting point or origin ("from"). | ['അമ്മ' സംഘടനയില് നിന്നും] → രാജിവച്ചു: "resigned from the 'Amma' organisation" |
| Goal | 13,171 | End point or destination ("to", "into"). | [ഇതിലേക്ക്] → ചേർക്കേണ്ടതാണ്: "should be added to this" |
| Temporal | 44,241 | When the event happens. | [ഭാവിയിൽ] ഫോണിന്റെ വില → വർദ്ധിക്കാൻ: "the phone's price may rise in future" |
| Manner | 28,870 | How the action is performed. | [പ്രകൃതിയുമായി ഇണങ്ങി] → ജീവിക്കുന്നത്: "living in harmony with nature" |
| Instrument | 6,301 | The tool or medium used. | [വെബ്സൈറ്റ് വഴി] അപേക്ഷ → സമർപ്പിക്കാവുന്നതാണ്: "can apply through the website" |
| Means | 3 | The method by which something is achieved. | (rare) |
| Degree | 4,770 | Extent or intensity. | [ഒരുപാട്] → വിഷമിപ്പിച്ചു: "upset (her) a lot" |
Logical relations: why, for what, under what condition, with what outcome
| Role | Count | Meaning | Example |
|---|---|---|---|
| Cause | 12,447 | The reason or cause of the event. | [ഈ സംഭവങ്ങളോടെ] ബന്ധം → അകന്നു: "the relationship drifted apart because of these events" |
| Purpose | 11,890 | The aim the action is meant to achieve. | [പ്രതിയെ പിടികൂടാൻ] തിരച്ചിൽ → ഊർജിതമാക്കി: "intensified the search to catch the accused" |
| Condition | 416 | The condition under which the event holds ("if"). | [വിഷമം തോന്നിയാൽ] → ഉറങ്ങാവുന്നതാണ്: "if you feel uneasy, you may sleep" |
| Result | 2,922 | The state or outcome the event produces. | സൗഹൃദം പിന്നീട് [പ്രണയമായി] → മാറി: "the friendship later turned into love" |
Why these roles matter
- Participant roles (Agent, Patient, Theme, Experiencer, Recipient…) answer who did what to whom. That's the core of an event, and of questions like "who resigned?" or "what was confirmed?".
- Circumstance roles (Location, Temporal, Manner, Source, Goal, Instrument) answer where, when and how. They're the backbone of event extraction and timeline building.
- Logical roles (Cause, Purpose, Condition, Result) capture why and with what effect. They're essential for reasoning, summarisation and causal question answering.
- In Malayalam these roles map only loosely onto surface form. The same dative suffix -ക്ക് can mark a Recipient, a Beneficiary or an Experiencer, and the same locative -ൽ can mark a Location or a Topic. Semantic roles make that distinction explicit, beyond what grammatical case alone shows.
Notes
- In a small number of very long sentences (0.4%), words near the end of the sentence carry no role.
- To read the TSV with pandas, turn off quote handling (
quoting=3); some words are a bare"character.
Files
| File | Description |
|---|---|
data/tokens.parquet |
tokens view (default config) |
data/frames.parquet |
frames view |
malayalam_semantic_tags.tsv |
tokens view as tab-separated text |
License
Released under the Creative Commons Attribution 4.0 International (CC BY 4.0) license. You are free to use, share and adapt this dataset, including commercially, with appropriate credit.
- Downloads last month
- 37